Upload mouse-run-run-2-transformer: trained agents, rollouts, evaluation, cross-architecture study

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+derived/reports/figures/evaluation_comparison.png filter=lfs diff=lfs merge=lfs -text
+derived/reports/figures/training_curves.png filter=lfs diff=lfs merge=lfs -text

HF_UPLOAD_MANIFEST.json

@@ -0,0 +1,674 @@
+{
+  "artifact_path": "runs/mouse-run-run-2-transformer",
+  "created_at": "2026-07-03T19:19:48.555969+00:00",
+  "excluded_from_file_manifest": [
+    "runs/mouse-run-run-2-transformer/HF_UPLOAD_MANIFEST.json"
+  ],
+  "file_count": 125,
+  "files": [
+    {
+      "path": "README.md",
+      "sha256": "02b919c0be5fbbfd2975aaa0accfdeb9350ecc300ef282256972ed5b2b569d0b",
+      "size_bytes": 2904
+    },
+    {
+      "path": "cross_architecture_study/cross_arch_cka.json",
+      "sha256": "1df526c3c10435203135df1ba02285eef813ebbaa176995178168313b6aeb797",
+      "size_bytes": 518
+    },
+    {
+      "path": "cross_architecture_study/cross_arch_perunit.json",
+      "sha256": "fda1247aa82af44396e5ab786e57622b4699290b6daf61319860442b72a4a852",
+      "size_bytes": 14851
+    },
+    {
+      "path": "cross_architecture_study/cross_arch_plsc_residualized.json",
+      "sha256": "596160749242ca86dee143e914c6c981654fa85bedb024f355c0a3236a909f70",
+      "size_bytes": 4062
+    },
+    {
+      "path": "cross_architecture_study/figures/cross_arch_cka.png",
+      "sha256": "7b43e1f229f0cec33101c1b89634082a5b84445e45e85ee114be19ad0c9c3dbb",
+      "size_bytes": 60221
+    },
+    {
+      "path": "cross_architecture_study/figures/sharing_vs_vision.png",
+      "sha256": "b9ddfbb26313e9013310fe5f7236dc50a5bc7ba4af2e8dff9c90331a64ad4f2a",
+      "size_bytes": 74823
+    },
+    {
+      "path": "cross_architecture_study/report.md",
+      "sha256": "26dde4d9e0348308c638aafa454f34404b3b81667fdfa24b24f67d47f015bf71",
+      "size_bytes": 11540
+    },
+    {
+      "path": "derived/reports/MANIFEST.json",
+      "sha256": "646d5089aeff3e68f08e4df142aae32fe3eaae03e4d6c7c5b1accc3b1751d913",
+      "size_bytes": 2782
+    },
+    {
+      "path": "derived/reports/figures/evaluation_comparison.png",
+      "sha256": "738e6a7547a9273692572f73c129ed2efb88ce00f18cfe1cd45f901a1235df4f",
+      "size_bytes": 111943
+    },
+    {
+      "path": "derived/reports/figures/training_curves.png",
+      "sha256": "9b4efbf32536bd83ad3bf2bfac46159b0cf15cc6d83acfd03820807246092357",
+      "size_bytes": 343071
+    },
+    {
+      "path": "derived/reports/report.md",
+      "sha256": "07ea6e1146a6d4391c08ed0d3763306a39917b879cc982c34946a4f2d9750cbd",
+      "size_bytes": 3017
+    },
+    {
+      "path": "derived/reports/summary.json",
+      "sha256": "7c1b08c438b29c8e14d9d9720f4cd9aa3e080f2aaa96b541b62351a9301a09d6",
+      "size_bytes": 14897
+    },
+    {
+      "path": "derived/tables/MANIFEST.json",
+      "sha256": "da05a5245daba06f81c15a1d2825867050fbb101cd80d8f0a879f95157bcf3a2",
+      "size_bytes": 2384
+    },
+    {
+      "path": "derived/tables/evaluations.jsonl",
+      "sha256": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f",
+      "size_bytes": 139093
+    },
+    {
+      "path": "derived/tables/exclusions.jsonl",
+      "sha256": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1",
+      "size_bytes": 3663
+    },
+    {
+      "path": "derived/tables/panels.json",
+      "sha256": "cb73e84a077cbcd2d639c1c7f6947f3a418fa7482b62d1b789d5ad9608d30562",
+      "size_bytes": 760
+    },
+    {
+      "path": "derived/tables/rollouts.jsonl",
+      "sha256": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b",
+      "size_bytes": 24314
+    },
+    {
+      "path": "derived/tables/runs.jsonl",
+      "sha256": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6",
+      "size_bytes": 50500
+    },
+    {
+      "path": "launch_gate.json",
+      "sha256": "0311afcdda4a0d9924e172fc0383e09f1a761124372d9e13999d37a60c2fe521",
+      "size_bytes": 4802
+    },
+    {
+      "path": "manifest.json",
+      "sha256": "2ac15af023b115bba05d12a98696623261af96c3cadd1936598c15d14869eeca",
+      "size_bytes": 15186
+    },
+    {
+      "path": "non_social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "bebcbee7c71f93593b985b9403dddac3f4b9d301088feea80adec365e5c839d3",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "non_social/seed_0000/attempt_01/config.json",
+      "sha256": "3886076e3107b0a4f5a93e3ad03817d35e2c0ce09d49a016d49bec2aa3877385",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0000/attempt_01/metrics.jsonl",
+      "sha256": "75912192efe4be72c477cc7de037d5e9df4a6e7a89cb9c75e9a4f965b1e51868",
+      "size_bytes": 54458
+    },
+    {
+      "path": "non_social/seed_0000/attempt_01/status.json",
+      "sha256": "8c66b9b3f5e2cb8af6d728d5486ca9b09a6c998bffd9cd2250871f31f134ec07",
+      "size_bytes": 1393
+    },
+    {
+      "path": "non_social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "97e55e51803fcb1f12344ef9426ee05617a44dd6420294c15ba2d71524feac2c",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "non_social/seed_0001/attempt_01/config.json",
+      "sha256": "d024907cb6c1ada7871a6aa3200527ec41df6b356ab615ef9be8e466e6c9f540",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0001/attempt_01/metrics.jsonl",
+      "sha256": "c77de502df9a63405ebf7483cf79c6a3b19f5956ae76bdb76b4e8ad731c7d08f",
+      "size_bytes": 54435
+    },
+    {
+      "path": "non_social/seed_0001/attempt_01/status.json",
+      "sha256": "db65d653e314dbb5f0c78a0d353dcda94e56c97f4fc3d12e1f56494f8c073028",
+      "size_bytes": 1389
+    },
+    {
+      "path": "non_social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "6b05e3f4ca6b8f514ca68a393b1bdc046876c2fdb9517fb51b9942361d3a57ac",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "non_social/seed_0002/attempt_01/config.json",
+      "sha256": "86dadf6a81b16adcd9d8c5b49178175a60848362f59fb2f4248aaaf8f0be73b8",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0002/attempt_01/metrics.jsonl",
+      "sha256": "abd38be484fbcf3287052a7e04ba148b97fbe5441c7760b7c2bc4ce10084fe3f",
+      "size_bytes": 54545
+    },
+    {
+      "path": "non_social/seed_0002/attempt_01/status.json",
+      "sha256": "3da5035da36a2c3aea567515cd5893a766b08e1fabe80389b8ff0bccccde1fbc",
+      "size_bytes": 1397
+    },
+    {
+      "path": "non_social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "a06bb3351d710015bd4527f14150baa0ed7be3ea4e957deb0bac2ad812840a89",
+      "size_bytes": 29760352
+    },
+    {
+      "path": "non_social/seed_0003/attempt_01/config.json",
+      "sha256": "58774ca21b63655b8b786e8d3aacc1d76f3fa7aef96888600c429bbe1d558c05",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0003/attempt_01/metrics.jsonl",
+      "sha256": "9efff90342274c9f9134bb9142940fef6fdf04431c413a5fe6ff3f0d6a3879aa",
+      "size_bytes": 54335
+    },
+    {
+      "path": "non_social/seed_0003/attempt_01/status.json",
+      "sha256": "ffd17a46f134f1144e851364b021ec18bf47a2178d033e289c563a817fb0644a",
+      "size_bytes": 1376
+    },
+    {
+      "path": "non_social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "7fc0051ad27e22e8e395faf9f846864d859748404592f14a524f3596f1f93559",
+      "size_bytes": 29760360
+    },
+    {
+      "path": "non_social/seed_0004/attempt_01/config.json",
+      "sha256": "2c87c1b048a5640816c623fa4dd03a45a5e69a1644a987d8d8c9b42df12c2153",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0004/attempt_01/metrics.jsonl",
+      "sha256": "dd2ff87c4c056c9b7babc70781047e108f6cbafeff2ad5b06ad6c8894c7eb881",
+      "size_bytes": 54469
+    },
+    {
+      "path": "non_social/seed_0004/attempt_01/status.json",
+      "sha256": "7a1c2a675f244d0d37c3379d7e7907ac4998fba181070e84dae6d2b2041b112c",
+      "size_bytes": 1382
+    },
+    {
+      "path": "non_social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "ac8c8ce4aab7af2d19fc15778486415942ca2cc3a405b65f8c5b23c7cfb69376",
+      "size_bytes": 29760360
+    },
+    {
+      "path": "non_social/seed_0005/attempt_01/config.json",
+      "sha256": "26fcacf461990235094e53773c91fcc8cf1f2495b4b8ced68e9088c7d91a1952",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0005/attempt_01/metrics.jsonl",
+      "sha256": "05f4bcab7222f1ad1aaa622e855d60c274bc2876a19a96c4bbec3fc65c00fcf8",
+      "size_bytes": 54666
+    },
+    {
+      "path": "non_social/seed_0005/attempt_01/status.json",
+      "sha256": "57db486373f69de367239a75f4d2327dd0ac14b2c6582982a675c6646e785dc1",
+      "size_bytes": 1384
+    },
+    {
+      "path": "non_social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "c3eb7fb2c91b2c029361c2a8a5a023ab29aba5b6cae2aca1cd33fac54dfe1d6a",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "non_social/seed_0006/attempt_01/config.json",
+      "sha256": "614faee0db30f57d9d2aee9fec5a10ba51bd15ad50b6ea500633615e240d412c",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0006/attempt_01/metrics.jsonl",
+      "sha256": "c376069bfc0b0097ee1a1dd3d193cfe9f5e78024861cf4fd4302903878ab1501",
+      "size_bytes": 54548
+    },
+    {
+      "path": "non_social/seed_0006/attempt_01/status.json",
+      "sha256": "135a01d6bb28e9a97e7737b39b2d80e6e8936bc6481951076ab918b4b384f3ec",
+      "size_bytes": 1394
+    },
+    {
+      "path": "non_social/seed_0007/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "021e19e4992a272e4a09a90a4bb4bbedf2703d0d2a2122af4b17a128246da307",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "non_social/seed_0007/attempt_01/config.json",
+      "sha256": "9bb3289524f274d1bf4efe1bbd94504490615579360be843e0307a9b6c07fbd4",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0007/attempt_01/metrics.jsonl",
+      "sha256": "5c14924f66958cffe3728c0c7ccd721dbbe68f842897f3bc501831bf655cdd45",
+      "size_bytes": 54589
+    },
+    {
+      "path": "non_social/seed_0007/attempt_01/status.json",
+      "sha256": "089bf762f103acfc6608f77b3d1bc982dcb4fc284a162e82ae0b26b1461d6529",
+      "size_bytes": 1396
+    },
+    {
+      "path": "non_social/seed_0008/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "8bd159d34635da80bf0bdbc948924eb30df6571f77a8dc53b6d138adad560333",
+      "size_bytes": 29760336
+    },
+    {
+      "path": "non_social/seed_0008/attempt_01/config.json",
+      "sha256": "d09109c22646e93c0c4dde6c32d7ecc6ec9155f8df6f422b3b77b828676f0a70",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0008/attempt_01/metrics.jsonl",
+      "sha256": "b1525ff219fe287b430e9b4c105641edd78090a90db29a24e538ee2c9122801e",
+      "size_bytes": 54465
+    },
+    {
+      "path": "non_social/seed_0008/attempt_01/status.json",
+      "sha256": "7c5f288999d30389b6cae79e66880601588e00105372d09fa125801512b6c023",
+      "size_bytes": 1361
+    },
+    {
+      "path": "non_social/seed_0009/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "6c3ce75ac26f6605968ebb1613d72e3f3e59ac898be34d14b34a59812bd25a49",
+      "size_bytes": 29760344
+    },
+    {
+      "path": "non_social/seed_0009/attempt_01/config.json",
+      "sha256": "b24d20d38c13df5a094b09cbdc16c8b4cba523d784c1baceff42c8210869e1dc",
+      "size_bytes": 1225
+    },
+    {
+      "path": "non_social/seed_0009/attempt_01/metrics.jsonl",
+      "sha256": "9fc047903def1bc2c246da61dc8a10ede51440f6654610a56edf41505e73532a",
+      "size_bytes": 54511
+    },
+    {
+      "path": "non_social/seed_0009/attempt_01/status.json",
+      "sha256": "fb931a5143914c38d6ab32d4a2a7ebedc1e565345fde56d044a02bc70b12c4d0",
+      "size_bytes": 1369
+    },
+    {
+      "path": "paper_random_eval.jsonl",
+      "sha256": "d78f6495e5f264d00f8b8bf084c877871b52a88d58d55efdb57de7219a6b5827",
+      "size_bytes": 135342
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors",
+      "sha256": "79523bf93d01ed70b7e868b54bd70459c210b81fefc7e06836c12becd787418b",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors",
+      "sha256": "4a525f188e61798e466bfab07962c00ba3a32419728ab58c132854d5c15021a6",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors",
+      "sha256": "1a3ffcae843364d6a41194d3fa074e3e38567f536a4c773bfce6d18933683747",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors",
+      "sha256": "fdff410f6f4486b5465b7d7e9b0eea876e81ec84939a13d65c18966e634c34f7",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors",
+      "sha256": "8a30011c7be774107fd2f2ec56a1411497995644a45e40ae9042e69fe31b8a7d",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors",
+      "sha256": "331de7d69aa252a1e6c25f85652611cfe8c65e76dc2d65502eea41dba4370686",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors",
+      "sha256": "49399764286403a10871d4a229bf4c66cc34b98d57dbf844493500003cff4e74",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors",
+      "sha256": "bdd9bfaf3009fd86f8b51becb610eac87ef0e29815d79f8a036fc7c78907441a",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors",
+      "sha256": "c46462eb90ada39824c0444c9e0be4903798ac5be414d9cf2d6edf098253bb45",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors",
+      "sha256": "2f95b021dd1d9ea44cf0be2a67663e3e14a847b79713c15072c335dd1ee53f51",
+      "size_bytes": 46506715
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors",
+      "sha256": "7b692d07d996c19666bcec9d3cbd1ca49895bbdaf265f987065c8c0a07a5bf67",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors",
+      "sha256": "5c982d42a21462cc56fbabd3d2e3cd0d0367cf46aad3e7751355630c4c16a448",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors",
+      "sha256": "fbd829a179d3930a2e2023754d754bfcd5da946d6a364fee2c4c15764e81cf5a",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors",
+      "sha256": "7850ba0b8693746b2fc891d045f690281838a55ff69f0ca4e7b0792ffb600089",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors",
+      "sha256": "4d30fd2cf0d2050b4e36b40486d11328ffb2f6223eb365ed5d24adebf7a042d6",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors",
+      "sha256": "61cab96377113bf461cc9b98496f068e46843299576259856285e4c5173396e4",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors",
+      "sha256": "fd6b3b52c19b03cbe1bc347b56180787b658f72ee1d28efeddae9a403e23ca1d",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors",
+      "sha256": "7ab97b49dbf08bc614e87382b28a0987ed96f4d5d3033b8b522164c2433de0fe",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors",
+      "sha256": "efa669fe4caed034302ce55a56e212b2cc51232ab9140ebeaf18711ad69dabb0",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors",
+      "sha256": "148864f819405af7554f128fa3574e5d75c0bb6afc957d717e24cb5752490e52",
+      "size_bytes": 46506699
+    },
+    {
+      "path": "paper_rollouts/paper_rollout_records.jsonl",
+      "sha256": "4d176c4e2f7a0290740ed4ca9122e1949522b03b9ef01cbf931c1f0b3c51186e",
+      "size_bytes": 21020
+    },
+    {
+      "path": "raw_records.jsonl",
+      "sha256": "445504330bfebe2c30711d6ef170368f4dae3cb3e677aa1e5d3babea4d64b75c",
+      "size_bytes": 47050
+    },
+    {
+      "path": "run_status.json",
+      "sha256": "230f229245a5e2d6608d281605d77ca32e9c0dcacb1e1da86fd1ebd862c0434f",
+      "size_bytes": 39468
+    },
+    {
+      "path": "social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "d43d74b10a625e6385b2bb95ea1788823a3302c478fc62063d73fa7275a2e64b",
+      "size_bytes": 29760376
+    },
+    {
+      "path": "social/seed_0000/attempt_01/config.json",
+      "sha256": "ac913cc7582a6ea11146af201067c0dd0ca6039bd301a218690bb9a758fb8964",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0000/attempt_01/metrics.jsonl",
+      "sha256": "94cf22269f9a2350ef0bed9daeb3318e24ce6e6145a6717f5d66f2eaea04df1a",
+      "size_bytes": 55395
+    },
+    {
+      "path": "social/seed_0000/attempt_01/status.json",
+      "sha256": "33b0214be2971b51f3db1c5e4beb4cdc23473c62b1a462d2a45e1dfeedb19f1e",
+      "size_bytes": 1405
+    },
+    {
+      "path": "social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "f65e21859c2db411274fb48eff4489b878589dc01d994df540c2fd45a29d7070",
+      "size_bytes": 29760384
+    },
+    {
+      "path": "social/seed_0001/attempt_01/config.json",
+      "sha256": "cdfbf77007da5b1e1056324dbec5d24a5c12fd430751c774c86e1481d1d46591",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0001/attempt_01/metrics.jsonl",
+      "sha256": "58bcc923be8e7341a9f6f333c05f80e2084e6e53a61db9aa0a2fe9bd663abe23",
+      "size_bytes": 55504
+    },
+    {
+      "path": "social/seed_0001/attempt_01/status.json",
+      "sha256": "cf5cfaba4653eeafec5b0d11eafda608a4ce5f6470e8afe3b701b2ae822a5261",
+      "size_bytes": 1412
+    },
+    {
+      "path": "social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "c1f4144d32c140593489ce5522c50515a469ce6c2ad5aae7e6125078757e8ada",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "social/seed_0002/attempt_01/config.json",
+      "sha256": "338e7d109047f61bec737052cf8b8c3debd4dad66d084e7954c6acb2b408eda3",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0002/attempt_01/metrics.jsonl",
+      "sha256": "e440afee038286e728ce19d5af9dccc9245dd0c9bc5c2a2a9e0b6a9ec3a71d89",
+      "size_bytes": 55488
+    },
+    {
+      "path": "social/seed_0002/attempt_01/status.json",
+      "sha256": "c0acbd148b7eef4c26f411b298bb22e8b18bafe0d0676eb9c96ab82a4ce3a5ee",
+      "size_bytes": 1402
+    },
+    {
+      "path": "social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "42e883b23fb099df0bbca6fff98d93505336beb9d24fcca6844a6a86809443c2",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "social/seed_0003/attempt_01/config.json",
+      "sha256": "1cd7b200d980d1f437677a0f8991dda2cd44ed80de15471f229978ff4b525cfd",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0003/attempt_01/metrics.jsonl",
+      "sha256": "18f56873b94f1b2fc72534fa3a30bf0a609123a225d66aa537eb3f363768bb02",
+      "size_bytes": 55429
+    },
+    {
+      "path": "social/seed_0003/attempt_01/status.json",
+      "sha256": "6c5d14007b34fe6cae16e2fea4787421ab9f57e03b294edb1b677126f4792958",
+      "size_bytes": 1402
+    },
+    {
+      "path": "social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "b8b54f8a46f42267d8050df70d059bed5fffe88df8f02d91f48faba250143759",
+      "size_bytes": 29760376
+    },
+    {
+      "path": "social/seed_0004/attempt_01/config.json",
+      "sha256": "17a097f5a6d0b4e9367511937201418d7678682f862ef8b14402fee30a5ff36b",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0004/attempt_01/metrics.jsonl",
+      "sha256": "01f78b3d1e1c5ee35bd544ec1e52ee4ab68821f35a137d5fca80ce8d1d32556c",
+      "size_bytes": 55539
+    },
+    {
+      "path": "social/seed_0004/attempt_01/status.json",
+      "sha256": "1d405e52a854336c00db8ea295214df34ede4d7c05485c83ea11e7dd647cb2d3",
+      "size_bytes": 1412
+    },
+    {
+      "path": "social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "78eaccb338b45f6a03f515f2cee3766c9898f3ac6218f1e0839ec5175bd738c0",
+      "size_bytes": 29760344
+    },
+    {
+      "path": "social/seed_0005/attempt_01/config.json",
+      "sha256": "d3e7cd3414bdfd674fc783ef961336fb3b354f83777630b5a45dac92ed308778",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0005/attempt_01/metrics.jsonl",
+      "sha256": "5986539aec56bfe2a4e152631cfe863bc2f3cc08d6a24480d0447fc1507b4022",
+      "size_bytes": 55334
+    },
+    {
+      "path": "social/seed_0005/attempt_01/status.json",
+      "sha256": "cb0e5e507cde74f70ea86364c004728496426b3af162ba3e6f80be9797e4c398",
+      "size_bytes": 1383
+    },
+    {
+      "path": "social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "23e8a61977d52ecaa6a6bca53b80271a5b31024659be373ca295584a0cf5e3fa",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "social/seed_0006/attempt_01/config.json",
+      "sha256": "ab1cd032b5fd5a4944b480e17fef6949b97e25f979aed5999c1f7098b0aa1dc5",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0006/attempt_01/metrics.jsonl",
+      "sha256": "a0ebcd73df03367aba224d567299249a175f00bec4632eb4da47c56107252638",
+      "size_bytes": 55460
+    },
+    {
+      "path": "social/seed_0006/attempt_01/status.json",
+      "sha256": "5bdb94a91f3f1e423934673f24dd47f6f9ff53fc76ad77055bf46717b673ad7c",
+      "size_bytes": 1403
+    },
+    {
+      "path": "social/seed_0007/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "31976e8959949de1246f21e9902b4882a9b47763a935458165086ad52ff66214",
+      "size_bytes": 29760376
+    },
+    {
+      "path": "social/seed_0007/attempt_01/config.json",
+      "sha256": "5514fc1f1cf4be651fd36a5027306de97959e429f343f3f0b1784918b6f6c040",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0007/attempt_01/metrics.jsonl",
+      "sha256": "160836bcdb07381ae668df54079a14604f1bbbb1b501068ebb12430c907aed44",
+      "size_bytes": 55456
+    },
+    {
+      "path": "social/seed_0007/attempt_01/status.json",
+      "sha256": "dafa1372eb70529ea83a90bcbe746a48bd2a400771c62dc1415d71f1f006a3c8",
+      "size_bytes": 1411
+    },
+    {
+      "path": "social/seed_0008/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "ff3f0e0f69d3eb842d30e1463bfb424b5f279fa52d10afb3cf854b7b57747b13",
+      "size_bytes": 29760384
+    },
+    {
+      "path": "social/seed_0008/attempt_01/config.json",
+      "sha256": "922725c0461eec84ba313fce07fc98c90ec2b09e5b4a4ced872b3560fe5fd828",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0008/attempt_01/metrics.jsonl",
+      "sha256": "82a71aa50ab724d0aa3f6427a14627ed360e81074e179d3a78b208a11c7b28f5",
+      "size_bytes": 55464
+    },
+    {
+      "path": "social/seed_0008/attempt_01/status.json",
+      "sha256": "c6d62b1b85f2e595a71c6e408c79ee37baf0533bc8d48000cdc91ee74c23a3e8",
+      "size_bytes": 1417
+    },
+    {
+      "path": "social/seed_0009/attempt_01/checkpoints/latest.safetensors",
+      "sha256": "d50b2b6c17101b2b486d954abdd7be83fe3678f1e69fd0fffeec0607f1fc7685",
+      "size_bytes": 29760368
+    },
+    {
+      "path": "social/seed_0009/attempt_01/config.json",
+      "sha256": "352697e7d6177a28a3e8281abc1ede3a27c3cdaae9dea03ef57906d7dc8ca40e",
+      "size_bytes": 1212
+    },
+    {
+      "path": "social/seed_0009/attempt_01/metrics.jsonl",
+      "sha256": "14993be488100bc62c64edad70f025737e9490b7761ec961bc7b99ad071e4f10",
+      "size_bytes": 55398
+    },
+    {
+      "path": "social/seed_0009/attempt_01/status.json",
+      "sha256": "19c1aacf7ae3caeba46581065da684a98425ee41d0e30b5632908a34938a1da0",
+      "size_bytes": 1401
+    },
+    {
+      "path": "triton_equivalence.json",
+      "sha256": "6c51db69737bc2cc7696492f10ad3c661f0f0add505c8b7153e9a609403d3ed2",
+      "size_bytes": 1951
+    }
+  ],
+  "path_in_repo": "",
+  "private": false,
+  "provenance": {
+    "command": "/Users/jacoblincool/Documents/GitHub/mouse-run-run/.venv/bin/upload-marl-artifacts runs/mouse-run-run-2-transformer --repo-id JacobLinCool/mouse-run-run-2-transformer --repo-type model --commit-message 'Upload mouse-run-run-2-transformer: trained agents, rollouts, evaluation, cross-architecture study'",
+    "created_at": "2026-07-03T19:19:49.136304+00:00",
+    "dependencies": {
+      "huggingface_hub": "1.21.0",
+      "numpy": "2.5.0",
+      "safetensors": "0.8.0",
+      "tensorboard": "2.21.0",
+      "torch": "2.12.1",
+      "triton": null
+    },
+    "executable": "/Users/jacoblincool/Documents/GitHub/mouse-run-run/.venv/bin/python3",
+    "git": {
+      "branch": "main",
+      "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "dirty": false,
+      "sha": "043425050d6ebb6201f0e8be035085405cb044ea",
+      "short_sha": "0434250",
+      "status_short": ""
+    },
+    "platform": "macOS-26.4-arm64-arm-64bit-Mach-O",
+    "python": "3.14.4 (main, Apr 14 2026, 14:46:33) [Clang 22.1.3 ]",
+    "schema_version": 1,
+    "torch": {
+      "cuda_available": false,
+      "cuda_device_count": 0,
+      "cuda_devices": [],
+      "cuda_version": null,
+      "cudnn_version": null,
+      "mps_available": true,
+      "version": "2.12.1"
+    }
+  },
+  "repo_id": "JacobLinCool/mouse-run-run-2-transformer",
+  "repo_type": "model",
+  "revision": null,
+  "schema_version": 1
+}

README.md

@@ -0,0 +1,71 @@
+---
+license: mit
+tags:
+- multi-agent-reinforcement-learning
+- ppo
+- recurrent-neural-network
+- neuroscience
+- inter-brain-dynamics
+---
+
+# mouse-run-run — TRANSFORMER agents (mouse-run-run-2-transformer)
+
+Trained chaser/explorer agents for a modern PyTorch reproduction and
+architecture extension of the MARL experiment in Zhang et al. (2025),
+*Inter-brain neural dynamics in biological and artificial intelligence
+systems*, Nature 645, 991–1001.
+
+**Architecture:** 2-layer causal transformer (no evolving state). Analysis (hidden) dimension 256; two independently
+parameterized actor-critic agents (chaser and explorer, no weight sharing).
+
+This repo holds one full experiment: **20 trained agent pairs** (tasks
+`social` and `non_social` × seeds 0–9), each trained with PPO for 20,000
+updates × 40 episodes × 100 steps. 20/20 units completed on the first attempt.
+
+## Contents
+
+- `{social,non_social}/seed_XXXX/attempt_01/checkpoints/latest.safetensors` —
+  trained chaser+explorer weights (safetensors; config/metrics in metadata).
+- `paper_rollouts/*.safetensors` — 25×500-step analysis rollouts per pair:
+  per-timestep hidden states, positions, actions, and social-event flags
+  (schema v3, time-aligned; degenerate episodes flagged).
+- `paper_random_eval.jsonl` — standardized random-opponent evaluation.
+- `derived/{tables,reports}/` — canonical tables and the per-experiment report.
+- `cross_architecture_study/` — the four-architecture comparison report,
+  figures, and cross-architecture PLSC/CKA data.
+- `manifest.json`, `raw_records.jsonl`, `launch_gate.json`,
+  `triton_equivalence.json` — provenance.
+
+## Key results (this architecture)
+
+Behavior (vs standardized random opponent, social task):
+- chaser collisions / episode: **41.31**
+  (non-social control: 3.39)
+- chaser partner-in-vision: 0.96
+
+Neural (social agents):
+- collision decoding (balanced acc): **0.85**
+- partner escape / approach decoding: 0.76 / 0.76
+- PLSC shared-dimension top correlation: 0.81
+
+See `cross_architecture_study/report.md` for the four-architecture comparison.
+**Headline:** the architectures do not learn the same internal representations —
+only the RNN develops genuine internal shared dynamics at low mutual vision.
+
+## Load a checkpoint
+
+```python
+from huggingface_hub import hf_hub_download
+from mouse_run_run.serialization import load_checkpoint
+from mouse_run_run.policy import build_policy
+
+path = hf_hub_download("JacobLinCool/mouse-run-run-2-transformer",
+    "social/seed_0000/attempt_01/checkpoints/latest.safetensors")
+config, metrics, chaser_state, explorer_state = load_checkpoint(path)
+chaser = build_policy(config["architecture"], config["env"]["observation_size"],
+                      hidden_size=config["hidden_size"])
+chaser.load_state_dict(chaser_state)
+```
+
+Code: https://github.com/JacobLinCool/mouse-run-run (paper reproduction +
+cross-architecture study).

cross_architecture_study/cross_arch_cka.json

@@ -0,0 +1,18 @@
+{
+ "RNN|RNN": 1.0,
+ "RNN|MLP": 0.6792286610675173,
+ "RNN|SSM": 0.6359547814617931,
+ "RNN|Transformer": 0.6660120097138259,
+ "MLP|RNN": 0.6792286610675173,
+ "MLP|MLP": 1.0,
+ "MLP|SSM": 0.7110754305522333,
+ "MLP|Transformer": 0.7809858975192786,
+ "SSM|RNN": 0.6359547814617931,
+ "SSM|MLP": 0.7110754305522333,
+ "SSM|SSM": 1.0,
+ "SSM|Transformer": 0.7551091733153037,
+ "Transformer|RNN": 0.6660120097138259,
+ "Transformer|MLP": 0.7809858975192786,
+ "Transformer|SSM": 0.7551091733153037,
+ "Transformer|Transformer": 1.0
+}

cross_architecture_study/cross_arch_perunit.json

@@ -0,0 +1,802 @@
+[
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 7.440000057220459,
+  "top_r": 0.03705123940656156
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 10.039999961853027,
+  "top_r": 0.09726206921336848
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 12.199999809265137,
+  "top_r": 0.09102936182631945
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 19.8799991607666,
+  "top_r": 0.1416026760369321
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 9.880000114440918,
+  "top_r": 0.06074593045795309
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 11.760000228881836,
+  "top_r": 0.06100869691148739
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 11.760000228881836,
+  "top_r": 0.047992243964107055
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 9.720000267028809,
+  "top_r": 0.040861022303547455
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 8.359999656677246,
+  "top_r": 0.046082531832419726
+ },
+ {
+  "arch": "RNN",
+  "task": "non_social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 15.119999885559082,
+  "top_r": 0.04697112454078715
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 16,
+  "vision": 0.16239520907402039,
+  "collisions": 60.91999816894531,
+  "top_r": 0.9606705270007906
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 18,
+  "vision": 0.21668662130832672,
+  "collisions": 0.800000011920929,
+  "top_r": 0.6658904391428773
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 22,
+  "vision": 0.3185628652572632,
+  "collisions": 5.519999980926514,
+  "top_r": 0.6149421028031533
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 3,
+  "status": "degenerate",
+  "n_valid": 1,
+  "vision": 0.24694611132144928,
+  "collisions": 47.279998779296875,
+  "top_r": null
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 23,
+  "vision": 0.9924950003623962,
+  "collisions": 150.9199981689453,
+  "top_r": 0.9469448413117497
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 23,
+  "vision": 0.03936127573251724,
+  "collisions": 0.5199999809265137,
+  "top_r": 0.24632233096607034
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.2863073945045471,
+  "collisions": 1.1200000047683716,
+  "top_r": 0.6195669473643807
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 7,
+  "status": "degenerate",
+  "n_valid": 2,
+  "vision": 0.2092614769935608,
+  "collisions": 21.84000015258789,
+  "top_r": null
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 14,
+  "vision": 0.13309381902217865,
+  "collisions": 19.959999084472656,
+  "top_r": 0.9244864792894123
+ },
+ {
+  "arch": "RNN",
+  "task": "social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.2818363308906555,
+  "collisions": 5.960000038146973,
+  "top_r": 0.5286762777215525
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 8.279999732971191,
+  "top_r": 0.07776644999751171
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 8.640000343322754,
+  "top_r": 0.07086241042694791
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 9.039999961853027,
+  "top_r": 0.10571891551983957
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 10.600000381469727,
+  "top_r": 0.12438404297121783
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 13.15999984741211,
+  "top_r": 0.08272535053372708
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 4.599999904632568,
+  "top_r": 0.05911174780800376
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 6.599999904632568,
+  "top_r": 0.12048705520440492
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 15.720000267028809,
+  "top_r": 0.20280221422494873
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 13.640000343322754,
+  "top_r": 0.1197033617321138
+ },
+ {
+  "arch": "MLP",
+  "task": "non_social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 11.920000076293945,
+  "top_r": 0.06501230641904866
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.981237530708313,
+  "collisions": 115.08000183105469,
+  "top_r": 0.68309682518684
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9638323187828064,
+  "collisions": 114.72000122070312,
+  "top_r": 0.6770425529610226
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9689421057701111,
+  "collisions": 105.16000366210938,
+  "top_r": 0.6294873925115205
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9881038069725037,
+  "collisions": 236.0800018310547,
+  "top_r": 0.7859041502514537
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9528942108154297,
+  "collisions": 125.0,
+  "top_r": 0.7282436596506869
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9758083820343018,
+  "collisions": 110.36000061035156,
+  "top_r": 0.6151424539297031
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9893013834953308,
+  "collisions": 114.23999786376953,
+  "top_r": 0.7531067455891379
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9883432984352112,
+  "collisions": 141.52000427246094,
+  "top_r": 0.6580303971938998
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9682235717773438,
+  "collisions": 101.19999694824219,
+  "top_r": 0.6171618473537643
+ },
+ {
+  "arch": "MLP",
+  "task": "social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.9904990196228027,
+  "collisions": 115.23999786376953,
+  "top_r": 0.639261743430404
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.0,
+  "collisions": 40.2400016784668,
+  "top_r": 0.28866651056915354
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 3.4800000190734863,
+  "top_r": 0.15885310687782797
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 3.799999952316284,
+  "top_r": 0.1087223462156361
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 1.8799999952316284,
+  "top_r": 0.0647609791042199
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 34.36000061035156,
+  "top_r": 0.6156083569490233
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 5,
+  "status": "degenerate",
+  "n_valid": 1,
+  "vision": 0.0,
+  "collisions": 34.68000030517578,
+  "top_r": null
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.0,
+  "collisions": 3.200000047683716,
+  "top_r": 0.08990168984095623
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 121.23999786376953,
+  "top_r": 0.6383316081199204
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 8,
+  "status": "degenerate",
+  "n_valid": 0,
+  "vision": 0.0,
+  "collisions": 46.31999969482422,
+  "top_r": null
+ },
+ {
+  "arch": "SSM",
+  "task": "non_social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 105.68000030517578,
+  "top_r": 0.5532722247875926
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 13,
+  "vision": 0.9386028051376343,
+  "collisions": 107.16000366210938,
+  "top_r": 0.6762547587766922
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 19,
+  "vision": 0.6916567087173462,
+  "collisions": 41.91999816894531,
+  "top_r": 0.5728661391129467
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 22,
+  "vision": 0.9957684874534607,
+  "collisions": 63.91999816894531,
+  "top_r": 0.5386769468869881
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 3,
+  "status": "degenerate",
+  "n_valid": 0,
+  "vision": 0.9954491257667542,
+  "collisions": 400.9200134277344,
+  "top_r": null
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 5,
+  "vision": 0.4332934021949768,
+  "collisions": 20.15999984741211,
+  "top_r": 0.6451686137949718
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.981157660484314,
+  "collisions": 66.23999786376953,
+  "top_r": 0.5719388049785209
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.9827544689178467,
+  "collisions": 63.79999923706055,
+  "top_r": 0.6399909552314786
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 8,
+  "vision": 0.9846706390380859,
+  "collisions": 129.55999755859375,
+  "top_r": 0.5951176365769322
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 3,
+  "vision": 0.8347305655479431,
+  "collisions": 139.27999877929688,
+  "top_r": 0.5911175381223174
+ },
+ {
+  "arch": "SSM",
+  "task": "social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 8,
+  "vision": 0.9699001908302307,
+  "collisions": 279.9599914550781,
+  "top_r": 0.9463852317330589
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.0,
+  "collisions": 18.15999984741211,
+  "top_r": 0.20624045007674752
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 177.52000427246094,
+  "top_r": 0.5466457412672473
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 247.67999267578125,
+  "top_r": 0.741521978186391
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 12,
+  "vision": 0.0,
+  "collisions": 32.47999954223633,
+  "top_r": 0.29652097824664553
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 89.83999633789062,
+  "top_r": 0.3499635714443735
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 118.91999816894531,
+  "top_r": 0.4220079232087302
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 15,
+  "vision": 0.0,
+  "collisions": 199.83999633789062,
+  "top_r": 0.4868188359531619
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 87.63999938964844,
+  "top_r": 0.3143182076125383
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 12.920000076293945,
+  "top_r": 0.10559527101392678
+ },
+ {
+  "arch": "Transformer",
+  "task": "non_social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.0,
+  "collisions": 4.920000076293945,
+  "top_r": 0.0921079793072239
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 0,
+  "status": "ok",
+  "n_valid": 23,
+  "vision": 0.8109381198883057,
+  "collisions": 66.27999877929688,
+  "top_r": 0.4658015424447193
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 1,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.816926121711731,
+  "collisions": 83.04000091552734,
+  "top_r": 0.37708898276609976
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 2,
+  "status": "ok",
+  "n_valid": 14,
+  "vision": 0.9682235717773438,
+  "collisions": 81.19999694824219,
+  "top_r": 0.4850167809908031
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 3,
+  "status": "ok",
+  "n_valid": 20,
+  "vision": 0.8345708847045898,
+  "collisions": 78.55999755859375,
+  "top_r": 0.4330689299750323
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 4,
+  "status": "ok",
+  "n_valid": 21,
+  "vision": 0.757125735282898,
+  "collisions": 76.16000366210938,
+  "top_r": 0.34081183250426467
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 5,
+  "status": "ok",
+  "n_valid": 20,
+  "vision": 0.6459081768989563,
+  "collisions": 100.72000122070312,
+  "top_r": 0.4682438954043613
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 6,
+  "status": "ok",
+  "n_valid": 25,
+  "vision": 0.7647106051445007,
+  "collisions": 47.84000015258789,
+  "top_r": 0.37878757497109683
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 7,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.9699800610542297,
+  "collisions": 95.5999984741211,
+  "top_r": 0.4278219769902787
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 8,
+  "status": "ok",
+  "n_valid": 20,
+  "vision": 0.7192016243934631,
+  "collisions": 103.23999786376953,
+  "top_r": 0.6244838845224036
+ },
+ {
+  "arch": "Transformer",
+  "task": "social",
+  "seed": 9,
+  "status": "ok",
+  "n_valid": 24,
+  "vision": 0.7394012212753296,
+  "collisions": 95.95999908447266,
+  "top_r": 0.3750948684332812
+ }
+]

cross_architecture_study/cross_arch_plsc_residualized.json

@@ -0,0 +1,401 @@
+{
+ "RNN": {
+  "social": [
+   [
+    0.9606705270007906,
+    156,
+    16
+   ],
+   [
+    0.6658904391428773,
+    236,
+    18
+   ],
+   [
+    0.6149421028031533,
+    250,
+    22
+   ],
+   [
+    0.9469448413117497,
+    256,
+    23
+   ],
+   [
+    0.24632233096607034,
+    190,
+    23
+   ],
+   [
+    0.6195669473643807,
+    255,
+    25
+   ],
+   [
+    0.9244864792894123,
+    212,
+    14
+   ],
+   [
+    0.5286762777215525,
+    256,
+    24
+   ]
+  ],
+  "non_social": [
+   [
+    0.03705123940656156,
+    54,
+    25
+   ],
+   [
+    0.09726206921336848,
+    111,
+    25
+   ],
+   [
+    0.09102936182631945,
+    108,
+    25
+   ],
+   [
+    0.1416026760369321,
+    116,
+    25
+   ],
+   [
+    0.06074593045795309,
+    95,
+    25
+   ],
+   [
+    0.06100869691148739,
+    122,
+    25
+   ],
+   [
+    0.047992243964107055,
+    101,
+    25
+   ],
+   [
+    0.040861022303547455,
+    113,
+    25
+   ],
+   [
+    0.046082531832419726,
+    120,
+    25
+   ],
+   [
+    0.04697112454078715,
+    121,
+    25
+   ]
+  ]
+ },
+ "MLP": {
+  "social": [
+   [
+    0.68309682518684,
+    256,
+    25
+   ],
+   [
+    0.6770425529610226,
+    256,
+    25
+   ],
+   [
+    0.6294873925115205,
+    256,
+    25
+   ],
+   [
+    0.7859041502514537,
+    256,
+    25
+   ],
+   [
+    0.7282436596506869,
+    256,
+    25
+   ],
+   [
+    0.6151424539297031,
+    256,
+    25
+   ],
+   [
+    0.7531067455891379,
+    256,
+    25
+   ],
+   [
+    0.6580303971938998,
+    256,
+    25
+   ],
+   [
+    0.6171618473537643,
+    256,
+    25
+   ],
+   [
+    0.639261743430404,
+    256,
+    25
+   ]
+  ],
+  "non_social": [
+   [
+    0.07776644999751171,
+    253,
+    25
+   ],
+   [
+    0.07086241042694791,
+    254,
+    25
+   ],
+   [
+    0.10571891551983957,
+    255,
+    25
+   ],
+   [
+    0.12438404297121783,
+    255,
+    25
+   ],
+   [
+    0.08272535053372708,
+    255,
+    25
+   ],
+   [
+    0.05911174780800376,
+    255,
+    25
+   ],
+   [
+    0.12048705520440492,
+    251,
+    25
+   ],
+   [
+    0.20280221422494873,
+    256,
+    25
+   ],
+   [
+    0.1197033617321138,
+    231,
+    25
+   ],
+   [
+    0.06501230641904866,
+    251,
+    25
+   ]
+  ]
+ },
+ "SSM": {
+  "social": [
+   [
+    0.6762547587766922,
+    253,
+    13
+   ],
+   [
+    0.5728661391129467,
+    256,
+    19
+   ],
+   [
+    0.5386769468869881,
+    249,
+    22
+   ],
+   [
+    0.6451686137949718,
+    249,
+    5
+   ],
+   [
+    0.5719388049785209,
+    256,
+    25
+   ],
+   [
+    0.6399909552314786,
+    256,
+    24
+   ],
+   [
+    0.5951176365769322,
+    239,
+    8
+   ],
+   [
+    0.5911175381223174,
+    179,
+    3
+   ],
+   [
+    0.9463852317330589,
+    224,
+    8
+   ]
+  ],
+  "non_social": [
+   [
+    0.28866651056915354,
+    235,
+    24
+   ],
+   [
+    0.15885310687782797,
+    137,
+    25
+   ],
+   [
+    0.1087223462156361,
+    99,
+    25
+   ],
+   [
+    0.0647609791042199,
+    185,
+    25
+   ],
+   [
+    0.6156083569490233,
+    244,
+    25
+   ],
+   [
+    0.08990168984095623,
+    100,
+    24
+   ],
+   [
+    0.6383316081199204,
+    252,
+    25
+   ],
+   [
+    0.5532722247875926,
+    235,
+    25
+   ]
+  ]
+ },
+ "Transformer": {
+  "social": [
+   [
+    0.4658015424447193,
+    251,
+    23
+   ],
+   [
+    0.37708898276609976,
+    251,
+    24
+   ],
+   [
+    0.4850167809908031,
+    253,
+    14
+   ],
+   [
+    0.4330689299750323,
+    254,
+    20
+   ],
+   [
+    0.34081183250426467,
+    249,
+    21
+   ],
+   [
+    0.4682438954043613,
+    241,
+    20
+   ],
+   [
+    0.37878757497109683,
+    250,
+    25
+   ],
+   [
+    0.4278219769902787,
+    249,
+    24
+   ],
+   [
+    0.6244838845224036,
+    249,
+    20
+   ],
+   [
+    0.3750948684332812,
+    248,
+    24
+   ]
+  ],
+  "non_social": [
+   [
+    0.20624045007674752,
+    20,
+    24
+   ],
+   [
+    0.5466457412672473,
+    152,
+    25
+   ],
+   [
+    0.741521978186391,
+    139,
+    25
+   ],
+   [
+    0.29652097824664553,
+    16,
+    12
+   ],
+   [
+    0.3499635714443735,
+    91,
+    25
+   ],
+   [
+    0.4220079232087302,
+    70,
+    25
+   ],
+   [
+    0.4868188359531619,
+    209,
+    15
+   ],
+   [
+    0.3143182076125383,
+    181,
+    25
+   ],
+   [
+    0.10559527101392678,
+    113,
+    25
+   ],
+   [
+    0.0921079793072239,
+    15,
+    25
+   ]
+  ]
+ }
+}

cross_architecture_study/report.md

@@ -0,0 +1,215 @@
+# Cross-Architecture Representation Study
+
+**Question.** Do different agent architectures, trained identically on the
+Zhang et al. (2025) chaser–explorer task, learn the *same* internal
+representations of the social interaction? Or does the phenomenon the paper
+reports — genuine shared neural dynamics between two independently trained
+agents — depend on the architecture?
+
+**Design.** Four architectures, identical in everything except the memory
+mechanism: reward structure, 10×10 world, 100-step episodes, seeds 0–9,
+analysis dimension 256, PPO settings, evaluation protocol, and analysis
+pipeline. They span a memory ladder:
+
+| architecture | memory mechanism | internal state at step t |
+|---|---|---|
+| MLP (`k=8` frame stack) | none (finite window) | pure function of the last 8 observations |
+| SSM (gated diagonal LRU) | compressed, **linear** dynamics | eigenvalue-controlled linear state |
+| RNN (paper's ReLU RNN) | compressed, **nonlinear** dynamics | chaotic nonlinear recurrence |
+| Transformer (2-layer causal) | none evolving; re-attends history | pre-head residual recomputed per step |
+
+Each was trained as a full 20-pair experiment (`mouse-run-run-1` for the RNN,
+`mouse-run-run-2-{mlp,ssm,transformer}` for the rest): 20,000 PPO updates ×
+40 episodes × 100 steps per pair, **20/20 units completed on the first
+attempt** for every architecture, zero failed/interrupted attempts.
+
+---
+
+## 1. Behavior: the architectures converged to different strategies
+
+Chaser capability against a standardized random explorer (100 episodes ×
+100 steps), and the explorer's learned escape (collisions it suffers against a
+random chaser — lower is better escape):
+
+| architecture | chaser collisions ↑ | chaser partner-in-vision | explorer collisions-suffered ↓ |
+|---|---:|---:|---:|
+| RNN | 8.98 ± 5.70 | 0.61 | 0.57 |
+| MLP | 26.93 ± 4.00 | 0.93 | 0.92 |
+| SSM | 35.92 ± 3.56 | 0.94 | 0.45 |
+| Transformer | 41.31 ± 3.29 | 0.96 | 0.28 |
+
+All four learned role-specific social behavior (non-social controls collide
+2.7–3.4× regardless). But the **RNN converged to a qualitatively different
+strategy**: a low-vision "stealth" chaser (61% partner-in-vision, 9 collisions)
+that tracks the explorer from memory, whereas MLP/SSM/Transformer converged to
+high-vision "pursuit" chasers (93–96% vision, 27–41 collisions) that stay
+glued to the explorer. This behavioral fork is the first sign the architectures
+are not learning the same thing — and it turns out to drive the representation
+difference.
+
+The raw capability ranking (Transformer > SSM > MLP > RNN) is **not** the
+headline. A chaser that collides 41×/episode by never leaving the explorer's
+side is behaviorally strong but representationally shallow. The scientific
+question is about the *internal representation*, below.
+
+## 2. Event decoding replicates across architectures (C2)
+
+Linear-classifier balanced accuracy for social events decoded from single-agent
+hidden states (5-fold CV, shuffled controls ≈ 0.50 throughout):
+
+| architecture | collision | partner escape | partner approach |
+|---|---:|---:|---:|
+| RNN | 0.983 | 0.888 | 0.886 |
+| MLP | 0.914 | 0.780 | 0.778 |
+| SSM | 0.850 | 0.768 | 0.733 |
+| Transformer | 0.847 | 0.761 | 0.759 |
+
+Every architecture encodes collision and partner approach/escape above chance
+in social agents — the paper's C2 claim is **architecture-robust**. Decoding
+accuracy tracks the memory ladder (RNN > MLP ≈ SSM > Transformer), but the
+qualitative result (social events are linearly represented) holds for all.
+
+## 3. Shared dimensions: a methodological correction, then the key result (C3)
+
+The paper's headline is C3: social agents develop *shared* neural dimensions
+across the two independently trained networks (PLSC). Measuring this across
+architectures exposed a confound that the analysis had to fix.
+
+**The raw cross-agent correlation is contaminated by time-locked scaffolds.**
+The Transformer's `top_dim_correlation` was **0.957 in the non-social control**
+— higher than its social value, and higher than any architecture's social
+score. There is no partner input in the non-social task, so this cannot be
+genuine sharing. The cause: the Transformer's positional embedding makes every
+timestep a deterministic function of absolute position t; both agents' networks
+carry this identical positional scaffold, and it dominates the top singular
+dimension. The RNN and MLP have no such clean scaffold (their "clocks" are noisy
+or absent), so their raw non-social correlation stays near zero — which is
+exactly why the raw statistic *looked* fine until the Transformer exposed it.
+
+**Fix (standard in neuroscience): remove the evoked/time-locked component.**
+Subtract the across-episode mean at each timestep — the response identical
+across episodes at a given t (positional embeddings, hidden-state clocks, mean
+approach dynamics) — and run PLSC on the episode-specific residual, where
+genuine interaction lives. This removes every architecture's deterministic
+temporal scaffold uniformly, making the effect size comparable.
+
+**Aggregate residualized top-dim correlation (interaction-driven sharing):**
+
+| architecture | social | non-social | separation |
+|---|---:|---:|---:|
+| RNN | 0.688 ± 0.23 | 0.067 | **0.62** |
+| MLP (floor) | 0.679 ± 0.06 | 0.103 | 0.58 |
+| SSM | 0.642 ± 0.12 | 0.315 | 0.33 |
+| Transformer | 0.438 ± 0.08 | 0.356 | **0.08** |
+
+The MLP is the **common-input floor**: being memoryless, any cross-agent sharing
+it shows is by construction reactive (both nets responding to the same visible
+game state). That its social sharing (0.679) nearly matches the RNN's aggregate
+(0.688) is the crux — the aggregate mean *hides* the real difference. The
+Transformer's social/non-social separation collapses to 0.08 once its positional
+scaffold is removed: **its apparent sharing was mostly the scaffold**.
+
+**The discriminating view: sharing vs behavioral coupling (vision).** The MLP
+can only share when both agents see each other. So the question is whether an
+architecture achieves shared representation *at low mutual vision* — which
+requires genuine internal (memory/prediction) coupling, not common input.
+Residualized social `top_r`, split by chaser partner-in-vision:
+
+| architecture | low vision (<40%) | high vision (≥40%) |
+|---|---:|---:|
+| **RNN** | **0.652 (n=7)** | 0.947 (n=1) |
+| MLP | — (0 units) | 0.679 (n=10) |
+| SSM | — (0 units) | 0.642 (n=9) |
+| Transformer | — (0 units) | 0.438 (n=10) |
+
+This is the result. **Only the RNN develops shared representation without the
+agents seeing each other** — 7 of its 8 analyzable social units sit below 40%
+mutual vision yet reach 0.65 residualized correlation. Every social unit of the
+other three architectures lives at high vision; their sharing is vision-gated,
+i.e. common-input-like. MLP *cannot* do otherwise (no memory). SSM and
+Transformer *have* memory mechanisms but did not use them to build low-vision
+coordination — they converged to the reactive, high-vision solution.
+
+## 4. Representational geometry: the RNN is the outlier (CKA)
+
+Feeding the **same** game-state observation stream through every architecture's
+social chaser and comparing representations with linear CKA (rotation/scaling
+invariant):
+
+|  | RNN | MLP | SSM | Transformer |
+|---|---:|---:|---:|---:|
+| RNN | 1.00 | 0.68 | 0.64 | 0.67 |
+| MLP | 0.68 | 1.00 | 0.71 | 0.78 |
+| SSM | 0.64 | 0.71 | 1.00 | 0.76 |
+| Transformer | 0.67 | 0.78 | 0.76 | 1.00 |
+
+The **RNN is the representational outlier** (0.64–0.68 with everything else),
+while MLP/SSM/Transformer cluster together (0.71–0.78). A third independent
+method agrees with behavior and PLSC: the RNN encodes the task differently; the
+other three converge to a more mutually similar (reactive) geometry.
+
+## 5. Degeneracy is itself an architecture property
+
+The paper excludes "degenerate" episodes where agents get stuck. The
+**SSM is markedly more prone to this**: 1/10 social and 2/10 non-social units
+were fully degenerate (too few valid episodes to analyze), and several more had
+only 3–8 of 25 episodes valid — versus **zero fully-degenerate units for RNN,
+MLP, and Transformer**. The linear recurrence appears to collapse toward
+fixed-point / stuck behavior more readily — a concrete failure mode of the
+linear-dynamics inductive bias on this task.
+
+---
+
+## Answer to the question
+
+**Different architectures do not learn the same internal representations.**
+Three independent methods converge:
+
+1. **Behavior** — the RNN alone converges to a low-vision, memory-based stealth
+   strategy; the others to high-vision reactive pursuit.
+2. **Shared dimensions (residualized, vision-conditioned)** — the RNN alone
+   develops genuine interaction-driven sharing when the agents *cannot see each
+   other*; the others' sharing is vision-gated and, for the Transformer, mostly
+   a positional scaffold once corrected.
+3. **CKA** — the RNN is the representational outlier; the other three cluster.
+
+The paper's hallmark phenomenon — genuine internal shared dynamics between
+independently trained agents — is **not architecture-universal**. It emerged
+cleanly only in the vanilla RNN. The memoryless MLP cannot produce it (its
+sharing is the common-input floor). The SSM and Transformer, despite having the
+capacity for memory, did not: their inductive biases steered them to a reactive,
+high-vision solution whose apparent sharing is largely explained by behavioral
+and input coupling, plus (Transformer) a deterministic temporal scaffold.
+
+## Honest confounds and limitations
+
+- **Strategy mediates representation.** The architectures converged to different
+  behavioral strategies, so the representation difference is partly *mediated* by
+  strategy, not a direct architecture→representation effect. But that is itself a
+  finding: the architecture's inductive bias shapes *which* solution is found,
+  and only the RNN's solution has genuine internal shared dynamics. Isolating a
+  direct effect would require conditioning on matched behavior (e.g. training all
+  architectures to the same vision regime), which these runs do not do.
+- **Not parameter-matched.** RNN 119k, SSM 382k, MLP 477k, Transformer 1.24M
+  params (analysis dim held at 256). The RNN's distinctness is not a
+  capacity-advantage artifact — it is the *smallest* model.
+- **Time-mean subtraction** removes any genuinely time-locked coordination along
+  with the scaffold; for random-initial-position episodes the interaction timing
+  is episode-specific, so this should be conservative, but it is an assumption.
+- **PLSC significant-dimension counts saturate** at these sample sizes (MLP hits
+  256/256 in both conditions); the effect size (residualized top correlation),
+  not the count, is the discriminating statistic and is what this report uses.
+- The **perturbation (C5)** and neural-action-space (C4) analyses are not yet run
+  across architectures; C5 in particular (does removing shared dimensions impair
+  social behavior) would test causality of the RNN's shared dimensions.
+
+## Artifacts
+
+- Per-experiment behavior/eval/rollouts/reports: `runs/{experiment}/`,
+  `runs/reports/{experiment}/`.
+- Neural analyses: `runs/analyses/{experiment}/neural_summary.json`.
+- Cross-architecture: `runs/analyses/cross_arch_plsc_residualized.json`,
+  `cross_arch_perunit.json`, `cross_arch_cka.json`.
+- Figures: `runs/analyses/figures/{sharing_vs_vision,cross_arch_cka}.png`.
+- Analysis code: `mouse_run_run/analysis.py`,
+  `scripts/analysis/neural_analyses.py`.

derived/reports/MANIFEST.json

@@ -0,0 +1,71 @@
+{
+  "analysis_script": "/root/mouse-run-run-perf/scripts/analysis/summarize_experiment.py",
+  "generated_at": "2026-07-03T01:24:40.542114+00:00",
+  "input_hashes": {
+    "evaluations.jsonl": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f",
+    "exclusions.jsonl": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1",
+    "rollouts.jsonl": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b",
+    "runs.jsonl": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6"
+  },
+  "input_tables": {
+    "evaluations.jsonl": "runs/tables/mouse-run-run-2-transformer/evaluations.jsonl",
+    "exclusions.jsonl": "runs/tables/mouse-run-run-2-transformer/exclusions.jsonl",
+    "rollouts.jsonl": "runs/tables/mouse-run-run-2-transformer/rollouts.jsonl",
+    "runs.jsonl": "runs/tables/mouse-run-run-2-transformer/runs.jsonl"
+  },
+  "output_hashes": {
+    "evaluation_comparison": "738e6a7547a9273692572f73c129ed2efb88ce00f18cfe1cd45f901a1235df4f",
+    "training_curves": "9b4efbf32536bd83ad3bf2bfac46159b0cf15cc6d83acfd03820807246092357"
+  },
+  "outputs": {
+    "figures": [
+      "runs/reports/mouse-run-run-2-transformer/figures/training_curves.png",
+      "runs/reports/mouse-run-run-2-transformer/figures/evaluation_comparison.png"
+    ],
+    "report": "runs/reports/mouse-run-run-2-transformer/report.md",
+    "summary": "runs/reports/mouse-run-run-2-transformer/summary.json"
+  },
+  "provenance": {
+    "command": "scripts/analysis/summarize_experiment.py runs/tables/mouse-run-run-2-transformer",
+    "created_at": "2026-07-03T01:24:40.542642+00:00",
+    "dependencies": {
+      "huggingface_hub": "1.21.0",
+      "numpy": "2.5.0",
+      "safetensors": "0.8.0",
+      "tensorboard": "2.21.0",
+      "torch": "2.12.1",
+      "triton": "3.7.1"
+    },
+    "executable": "/root/mouse-run-run-perf/.venv/bin/python3",
+    "git": {
+      "branch": "main",
+      "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "dirty": false,
+      "sha": "6282c650d281b31d646d537dd3f57c968806a423",
+      "short_sha": "6282c65",
+      "status_short": ""
+    },
+    "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39",
+    "python": "3.14.5 (main, Jun  2 2026, 22:25:17) [Clang 22.1.3 ]",
+    "schema_version": 1,
+    "torch": {
+      "cuda_available": true,
+      "cuda_device_count": 1,
+      "cuda_devices": [
+        {
+          "index": 0,
+          "major": 8,
+          "minor": 9,
+          "multi_processor_count": 66,
+          "name": "NVIDIA GeForce RTX 4070 Ti SUPER",
+          "total_memory_bytes": 16717578240
+        }
+      ],
+      "cuda_version": "13.0",
+      "cudnn_version": 92000,
+      "mps_available": false,
+      "version": "2.12.1+cu130"
+    }
+  },
+  "schema_version": 2
+}

derived/reports/report.md

@@ -0,0 +1,88 @@
+# MARL Experiment Summary
+
+Generated: `2026-07-03T01:24:40.539994+00:00`
+Tables: `runs/tables/mouse-run-run-2-transformer`
+
+## Valid Pairs
+
+| Task | Valid pairs |
+| --- | ---: |
+| `non_social` | 10 |
+| `social` | 10 |
+
+## Training Metrics (final update, mean ± std across pairs)
+
+| Metric | `non_social` (n=10) | `social` (n=10) |
+| --- | ---: | ---: |
+| collisions per episode | 0.748 ± 0.267 | 10.745 ± 2.13 |
+| chaser return | 8.787 ± 0.193 | 12.937 ± 2.166 |
+| explorer return | 94.041 ± 1.533 | 41.021 ± 5.261 |
+| chaser partner vision | 0 ± 0 | 0.97 ± 0.01 |
+| explorer partner vision | 0 ± 0 | 0.97 ± 0.01 |
+| chaser new fields | 94.165 ± 0.998 | 58.885 ± 2.309 |
+| explorer new fields | 96.328 ± 1.137 | 66.545 ± 2.836 |
+| final distance | 3.248 ± 0.764 | 1.673 ± 0.117 |
+| policy loss | -0.000632 ± 0.000994 | -0.005 ± 0.001 |
+| value loss | 0.08 ± 0.047 | 2.879 ± 0.469 |
+| entropy | 0.002 ± 0.000794 | 0.154 ± 0.027 |
+| approx kl | 0.000827 ± 0.001 | 0.003 ± 0.001 |
+| chaser grad norm | 0.058 ± 0.06 | 0.704 ± 0.362 |
+| explorer grad norm | 0.311 ± 0.15 | 0.957 ± 0.34 |
+| max param abs | 2.849 ± 0.52 | 2.752 ± 0.159 |
+
+## Random-Opponent Evaluation
+
+### Opponent: `random_chaser`
+
+| Metric | `non_social` (n=10) | `social` (n=10) |
+| --- | ---: | ---: |
+| collisions per episode | 2.797 ± 0.255 | 0.277 ± 0.052 |
+| chaser return | -3.25 ± 0.046 | -2.903 ± 0.061 |
+| explorer return | 90.611 ± 1.929 | 6.096 ± 7.816 |
+| chaser partner vision | 0 ± 0 | 0.163 ± 0.021 |
+| explorer partner vision | 0 ± 0 | 0.163 ± 0.021 |
+| chaser new fields | 34.021 ± 0.216 | 34.03 ± 0.058 |
+| explorer new fields | 94.393 ± 1.267 | 37.601 ± 5.236 |
+| final distance | 4.085 ± 0.228 | 6.707 ± 0.182 |
+| average distance | 5.08 ± 0.029 | 6.528 ± 0.209 |
+| degenerate fraction | 0.879 ± 0.018 | 0.96 ± 0.02 |
+
+### Opponent: `random_explorer`
+
+| Metric | `non_social` (n=10) | `social` (n=10) |
+| --- | ---: | ---: |
+| collisions per episode | 3.395 ± 0.471 | 41.308 ± 3.288 |
+| chaser return | 8.349 ± 0.203 | 40.052 ± 3.481 |
+| explorer return | -1.149 ± 0.219 | -43.318 ± 3.18 |
+| chaser partner vision | 0 ± 0 | 0.961 ± 0.007 |
+| explorer partner vision | 0 ± 0 | 0.961 ± 0.007 |
+| chaser new fields | 92.85 ± 0.907 | 26.842 ± 1.072 |
+| explorer new fields | 33.097 ± 0.088 | 23.326 ± 1.153 |
+| final distance | 4.388 ± 0.329 | 1.403 ± 0.083 |
+| average distance | 5.111 ± 0.029 | 1.571 ± 0.052 |
+| degenerate fraction | 0.853 ± 0.014 | 0.982 ± 0.015 |
+
+Failed evaluations: 0
+
+## Figures
+
+### training curves
+
+![training_curves](figures/training_curves.png)
+
+Training curves, mean ± std across 20 valid pairs.
+
+### evaluation comparison
+
+![evaluation_comparison](figures/evaluation_comparison.png)
+
+Evaluation metrics over 40 successful evaluations, grouped by opponent mode and task.
+
+## Degenerate Episode Exclusions
+
+| Excluded episodes | Episodes with records | Records |
+| ---: | ---: | ---: |
+| 59 | 300 | 12 |
+
+## Failed Attempts
+- count: 0

derived/reports/summary.json

@@ -0,0 +1,544 @@
+{
+  "evaluation_metrics": {
+    "failed_count": 0,
+    "non_social": {
+      "random_chaser": {
+        "average_distance": {
+          "max": 5.129418849945068,
+          "mean": 5.0796972751617435,
+          "min": 5.043961048126221,
+          "n": 10,
+          "std": 0.02873995140186124
+        },
+        "chaser_new_fields": {
+          "max": 34.36000061035156,
+          "mean": 34.020999908447266,
+          "min": 33.72999954223633,
+          "n": 10,
+          "std": 0.21644684003085007
+        },
+        "chaser_partner_vision": {
+          "max": 0.0,
+          "mean": 0.0,
+          "min": 0.0,
+          "n": 10,
+          "std": 0.0
+        },
+        "chaser_return": {
+          "max": -3.1859986782073975,
+          "mean": -3.2501988649368285,
+          "min": -3.3139989376068115,
+          "n": 10,
+          "std": 0.04577726762110156
+        },
+        "collisions_per_episode": {
+          "max": 3.059999942779541,
+          "mean": 2.796999955177307,
+          "min": 2.319999933242798,
+          "n": 10,
+          "std": 0.2553840545416018
+        },
+        "degenerate_fraction": {
+          "max": 0.91,
+          "mean": 0.8789999999999999,
+          "min": 0.86,
+          "n": 10,
+          "std": 0.01813835714721707
+        },
+        "explorer_new_fields": {
+          "max": 96.38999938964844,
+          "mean": 94.39299774169922,
+          "min": 92.62999725341797,
+          "n": 10,
+          "std": 1.267440183318429
+        },
+        "explorer_partner_vision": {
+          "max": 0.0,
+          "mean": 0.0,
+          "min": 0.0,
+          "n": 10,
+          "std": 0.0
+        },
+        "explorer_return": {
+          "max": 93.50499725341797,
+          "mean": 90.61149673461914,
+          "min": 87.93999481201172,
+          "n": 10,
+          "std": 1.9294984069099073
+        },
+        "final_distance": {
+          "max": 4.635434627532959,
+          "mean": 4.085444545745849,
+          "min": 3.837200880050659,
+          "n": 10,
+          "std": 0.22845185221708492
+        },
+        "n": 10
+      },
+      "random_explorer": {
+        "average_distance": {
+          "max": 5.154397487640381,
+          "mean": 5.11141185760498,
+          "min": 5.0569047927856445,
+          "n": 10,
+          "std": 0.028723257963978274
+        },
+        "chaser_new_fields": {
+          "max": 93.91999816894531,
+          "mean": 92.8499969482422,
+          "min": 90.43000030517578,
+          "n": 10,
+          "std": 0.9074680661154696
+        },
+        "chaser_partner_vision": {
+          "max": 0.0,
+          "mean": 0.0,
+          "min": 0.0,
+          "n": 10,
+          "std": 0.0
+        },
+        "chaser_return": {
+          "max": 8.579997062683105,
+          "mean": 8.349396276474,
+          "min": 7.805994510650635,
+          "n": 10,
+          "std": 0.20263148293859806
+        },
+        "collisions_per_episode": {
+          "max": 4.210000038146973,
+          "mean": 3.394999933242798,
+          "min": 2.6599998474121094,
+          "n": 10,
+          "std": 0.4708556415010507
+        },
+        "degenerate_fraction": {
+          "max": 0.87,
+          "mean": 0.853,
+          "min": 0.82,
+          "n": 10,
+          "std": 0.014177446878757837
+        },
+        "explorer_new_fields": {
+          "max": 33.21999740600586,
+          "mean": 33.09699897766113,
+          "min": 32.939998626708984,
+          "n": 10,
+          "std": 0.08809638710169938
+        },
+        "explorer_partner_vision": {
+          "max": 0.0,
+          "mean": 0.0,
+          "min": 0.0,
+          "n": 10,
+          "std": 0.0
+        },
+        "explorer_return": {
+          "max": -0.8899999856948853,
+          "mean": -1.149499958753586,
+          "min": -1.59499990940094,
+          "n": 10,
+          "std": 0.21923101569002812
+        },
+        "final_distance": {
+          "max": 5.0355939865112305,
+          "mean": 4.388345527648926,
+          "min": 4.020629405975342,
+          "n": 10,
+          "std": 0.3294317958141699
+        },
+        "n": 10
+      }
+    },
+    "social": {
+      "random_chaser": {
+        "average_distance": {
+          "max": 6.819681167602539,
+          "mean": 6.528078365325928,
+          "min": 6.161418914794922,
+          "n": 10,
+          "std": 0.20855749231374748
+        },
+        "chaser_new_fields": {
+          "max": 34.11000061035156,
+          "mean": 34.029999542236325,
+          "min": 33.939998626708984,
+          "n": 10,
+          "std": 0.057620230332741544
+        },
+        "chaser_partner_vision": {
+          "max": 0.20059996843338013,
+          "mean": 0.163489992916584,
+          "min": 0.13729999959468842,
+          "n": 10,
+          "std": 0.021343070859131442
+        },
+        "chaser_return": {
+          "max": -2.780999183654785,
+          "mean": -2.9034990072250366,
+          "min": -2.978998899459839,
+          "n": 10,
+          "std": 0.06094620981282601
+        },
+        "collisions_per_episode": {
+          "max": 0.3700000047683716,
+          "mean": 0.27699999809265136,
+          "min": 0.20999999344348907,
+          "n": 10,
+          "std": 0.05158488367146182
+        },
+        "degenerate_fraction": {
+          "max": 1.0,
+          "mean": 0.96,
+          "min": 0.93,
+          "n": 10,
+          "std": 0.020493901531919198
+        },
+        "explorer_new_fields": {
+          "max": 46.5,
+          "mean": 37.600999450683595,
+          "min": 30.35999870300293,
+          "n": 10,
+          "std": 5.235516137790574
+        },
+        "explorer_partner_vision": {
+          "max": 0.20059996843338013,
+          "mean": 0.163489992916584,
+          "min": 0.13729999959468842,
+          "n": 10,
+          "std": 0.021343070859131442
+        },
+        "explorer_return": {
+          "max": 19.404998779296875,
+          "mean": 6.096499764919281,
+          "min": -4.670000076293945,
+          "n": 10,
+          "std": 7.816309965617632
+        },
+        "final_distance": {
+          "max": 6.965867042541504,
+          "mean": 6.706823682785034,
+          "min": 6.317955017089844,
+          "n": 10,
+          "std": 0.18157252453346814
+        },
+        "n": 10
+      },
+      "random_explorer": {
+        "average_distance": {
+          "max": 1.667794942855835,
+          "mean": 1.5712934255599975,
+          "min": 1.4947947263717651,
+          "n": 10,
+          "std": 0.05184039584079168
+        },
+        "chaser_new_fields": {
+          "max": 28.6299991607666,
+          "mean": 26.841999435424803,
+          "min": 25.19999885559082,
+          "n": 10,
+          "std": 1.0724253811717708
+        },
+        "chaser_partner_vision": {
+          "max": 0.9711999893188477,
+          "mean": 0.9609299838542938,
+          "min": 0.9440000057220459,
+          "n": 10,
+          "std": 0.006767867998884705
+        },
+        "chaser_return": {
+          "max": 44.46000289916992,
+          "mean": 40.052001571655275,
+          "min": 34.46500015258789,
+          "n": 10,
+          "std": 3.4813239901279833
+        },
+        "collisions_per_episode": {
+          "max": 45.36000061035156,
+          "mean": 41.30799903869629,
+          "min": 36.189998626708984,
+          "n": 10,
+          "std": 3.2882451909664763
+        },
+        "degenerate_fraction": {
+          "max": 1.0,
+          "mean": 0.982,
+          "min": 0.96,
+          "n": 10,
+          "std": 0.01536229149573723
+        },
+        "explorer_new_fields": {
+          "max": 25.18000030517578,
+          "mean": 23.325999641418456,
+          "min": 21.5,
+          "n": 10,
+          "std": 1.1533272603517895
+        },
+        "explorer_partner_vision": {
+          "max": 0.9711999893188477,
+          "mean": 0.9609299838542938,
+          "min": 0.9440000057220459,
+          "n": 10,
+          "std": 0.006767867998884705
+        },
+        "explorer_return": {
+          "max": -38.09000015258789,
+          "mean": -43.3179988861084,
+          "min": -46.93000030517578,
+          "n": 10,
+          "std": 3.1799977794472576
+        },
+        "final_distance": {
+          "max": 1.5478826761245728,
+          "mean": 1.403444504737854,
+          "min": 1.2944831848144531,
+          "n": 10,
+          "std": 0.08294436722867861
+        },
+        "n": 10
+      }
+    }
+  },
+  "failed_attempts": [],
+  "figures": [
+    {
+      "caption": "Training curves, mean \u00b1 std across 20 valid pairs.",
+      "name": "training_curves",
+      "path": "runs/reports/mouse-run-run-2-transformer/figures/training_curves.png"
+    },
+    {
+      "caption": "Evaluation metrics over 40 successful evaluations, grouped by opponent mode and task.",
+      "name": "evaluation_comparison",
+      "path": "runs/reports/mouse-run-run-2-transformer/figures/evaluation_comparison.png"
+    }
+  ],
+  "generated_at": "2026-07-03T01:24:40.539994+00:00",
+  "rollout_exclusions": {
+    "episodes_with_exclusion_records": 300,
+    "excluded_degenerate_episodes": 59,
+    "records": 12
+  },
+  "schema_version": 2,
+  "tables_root": "runs/tables/mouse-run-run-2-transformer",
+  "training_metrics": {
+    "non_social": {
+      "approx_kl": {
+        "max": 0.0031268233263772593,
+        "mean": 0.0008274657820166141,
+        "min": 3.0426282704221208e-05,
+        "n": 10,
+        "std": 0.0011119185808850796
+      },
+      "chaser_grad_norm": {
+        "max": 0.2260504849255085,
+        "mean": 0.05761161218397319,
+        "min": 0.02223971695639193,
+        "n": 10,
+        "std": 0.06012663616499861
+      },
+      "chaser_new_fields": {
+        "max": 95.80000305175781,
+        "mean": 94.1650016784668,
+        "min": 91.82500457763672,
+        "n": 10,
+        "std": 0.9980102893568585
+      },
+      "chaser_partner_vision": {
+        "max": 0.0,
+        "mean": 0.0,
+        "min": 0.0,
+        "n": 10,
+        "std": 0.0
+      },
+      "chaser_return": {
+        "max": 9.135000228881836,
+        "mean": 8.787000560760498,
+        "min": 8.360000610351562,
+        "n": 10,
+        "std": 0.19291447540213363
+      },
+      "collisions_per_episode": {
+        "max": 1.149999976158142,
+        "mean": 0.7475000083446502,
+        "min": 0.3499999940395355,
+        "n": 10,
+        "std": 0.2667981449728105
+      },
+      "entropy": {
+        "max": 0.003422852954827249,
+        "mean": 0.0021465113706653936,
+        "min": 0.0009539475431665778,
+        "n": 10,
+        "std": 0.0007943372788127037
+      },
+      "explorer_grad_norm": {
+        "max": 0.6812433376908302,
+        "mean": 0.31067706923931837,
+        "min": 0.1489477176219225,
+        "n": 10,
+        "std": 0.15010061774843952
+      },
+      "explorer_new_fields": {
+        "max": 97.95000457763672,
+        "mean": 96.32750244140625,
+        "min": 94.82500457763672,
+        "n": 10,
+        "std": 1.1369062620540298
+      },
+      "explorer_partner_vision": {
+        "max": 0.0,
+        "mean": 0.0,
+        "min": 0.0,
+        "n": 10,
+        "std": 0.0
+      },
+      "explorer_return": {
+        "max": 96.0625,
+        "mean": 94.04125213623047,
+        "min": 91.82500457763672,
+        "n": 10,
+        "std": 1.5333274673425037
+      },
+      "final_distance": {
+        "max": 4.808191776275635,
+        "mean": 3.247923755645752,
+        "min": 2.2879505157470703,
+        "n": 10,
+        "std": 0.7642974377711885
+      },
+      "max_param_abs": {
+        "max": 3.969390392303467,
+        "mean": 2.849128818511963,
+        "min": 2.185072422027588,
+        "n": 10,
+        "std": 0.5202608551067897
+      },
+      "n": 10,
+      "policy_loss": {
+        "max": -4.3445531147767724e-05,
+        "mean": -0.0006323951178690334,
+        "min": -0.0035095140663869984,
+        "n": 10,
+        "std": 0.0009944805630664182
+      },
+      "value_loss": {
+        "max": 0.2153652273118496,
+        "mean": 0.08045755904167891,
+        "min": 0.040869749151170254,
+        "n": 10,
+        "std": 0.04654358081243757
+      }
+    },
+    "social": {
+      "approx_kl": {
+        "max": 0.004598107310812227,
+        "mean": 0.00303017191217893,
+        "min": 0.0012853747145520344,
+        "n": 10,
+        "std": 0.0010207158969125706
+      },
+      "chaser_grad_norm": {
+        "max": 1.5083578824996948,
+        "mean": 0.7038940638303757,
+        "min": 0.36728256940841675,
+        "n": 10,
+        "std": 0.3619897429634022
+      },
+      "chaser_new_fields": {
+        "max": 63.22500228881836,
+        "mean": 58.885000991821286,
+        "min": 55.75,
+        "n": 10,
+        "std": 2.30868508603447
+      },
+      "chaser_partner_vision": {
+        "max": 0.986750066280365,
+        "mean": 0.9699750483036041,
+        "min": 0.9477500319480896,
+        "n": 10,
+        "std": 0.010192195843222708
+      },
+      "chaser_return": {
+        "max": 16.552501678466797,
+        "mean": 12.937000370025634,
+        "min": 9.160000801086426,
+        "n": 10,
+        "std": 2.1664790907335263
+      },
+      "collisions_per_episode": {
+        "max": 14.475000381469727,
+        "mean": 10.745000123977661,
+        "min": 7.550000190734863,
+        "n": 10,
+        "std": 2.1296654790690113
+      },
+      "entropy": {
+        "max": 0.20668550580739975,
+        "mean": 0.15412018559873103,
+        "min": 0.11317355558276176,
+        "n": 10,
+        "std": 0.027327137994586925
+      },
+      "explorer_grad_norm": {
+        "max": 1.6371502578258514,
+        "mean": 0.9565056964755059,
+        "min": 0.60902339220047,
+        "n": 10,
+        "std": 0.3402408819474271
+      },
+      "explorer_new_fields": {
+        "max": 70.80000305175781,
+        "mean": 66.54500160217285,
+        "min": 62.5,
+        "n": 10,
+        "std": 2.835794556837627
+      },
+      "explorer_partner_vision": {
+        "max": 0.986750066280365,
+        "mean": 0.9699750483036041,
+        "min": 0.9477500319480896,
+        "n": 10,
+        "std": 0.010192195843222708
+      },
+      "explorer_return": {
+        "max": 47.42499923706055,
+        "mean": 41.021250534057614,
+        "min": 33.775001525878906,
+        "n": 10,
+        "std": 5.261294318837594
+      },
+      "final_distance": {
+        "max": 1.8629791736602783,
+        "mean": 1.6734241247177124,
+        "min": 1.4801658391952515,
+        "n": 10,
+        "std": 0.11725492127768533
+      },
+      "max_param_abs": {
+        "max": 2.9420251846313477,
+        "mean": 2.7518649816513063,
+        "min": 2.4855704307556152,
+        "n": 10,
+        "std": 0.15851267210951459
+      },
+      "n": 10,
+      "policy_loss": {
+        "max": -0.002069776682645852,
+        "mean": -0.004644430192469429,
+        "min": -0.007046535808569843,
+        "n": 10,
+        "std": 0.001406250863722305
+      },
+      "value_loss": {
+        "max": 3.7607646584510803,
+        "mean": 2.8793665051460264,
+        "min": 1.877354085445404,
+        "n": 10,
+        "std": 0.4693238146909516
+      }
+    }
+  },
+  "valid_pairs_by_task": {
+    "non_social": 10,
+    "social": 10
+  }
+}

derived/tables/MANIFEST.json

@@ -0,0 +1,67 @@
+{
+  "experiment_root": "runs/mouse-run-run-2-transformer",
+  "generated_at": "2026-07-03T01:24:38.539082+00:00",
+  "output_hashes": {
+    "evaluations": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f",
+    "exclusions": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1",
+    "rollouts": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b",
+    "runs": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6"
+  },
+  "outputs": {
+    "evaluations": "runs/tables/mouse-run-run-2-transformer/evaluations.jsonl",
+    "exclusions": "runs/tables/mouse-run-run-2-transformer/exclusions.jsonl",
+    "rollouts": "runs/tables/mouse-run-run-2-transformer/rollouts.jsonl",
+    "runs": "runs/tables/mouse-run-run-2-transformer/runs.jsonl"
+  },
+  "panel_path": "runs/tables/mouse-run-run-2-transformer/panels.json",
+  "provenance": {
+    "command": "scripts/transform/build_tables.py runs/mouse-run-run-2-transformer",
+    "created_at": "2026-07-03T01:24:38.539388+00:00",
+    "dependencies": {
+      "huggingface_hub": "1.21.0",
+      "numpy": "2.5.0",
+      "safetensors": "0.8.0",
+      "tensorboard": "2.21.0",
+      "torch": "2.12.1",
+      "triton": "3.7.1"
+    },
+    "executable": "/root/mouse-run-run-perf/.venv/bin/python3",
+    "git": {
+      "branch": "main",
+      "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "dirty": false,
+      "sha": "6282c650d281b31d646d537dd3f57c968806a423",
+      "short_sha": "6282c65",
+      "status_short": ""
+    },
+    "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39",
+    "python": "3.14.5 (main, Jun  2 2026, 22:25:17) [Clang 22.1.3 ]",
+    "schema_version": 1,
+    "torch": {
+      "cuda_available": true,
+      "cuda_device_count": 1,
+      "cuda_devices": [
+        {
+          "index": 0,
+          "major": 8,
+          "minor": 9,
+          "multi_processor_count": 66,
+          "name": "NVIDIA GeForce RTX 4070 Ti SUPER",
+          "total_memory_bytes": 16717578240
+        }
+      ],
+      "cuda_version": "13.0",
+      "cudnn_version": 92000,
+      "mps_available": false,
+      "version": "2.12.1+cu130"
+    }
+  },
+  "row_counts": {
+    "evaluations": 40,
+    "exclusions": 12,
+    "rollouts": 20,
+    "runs": 20
+  },
+  "schema_version": 1,
+  "transform_script": "/root/mouse-run-run-perf/scripts/transform/build_tables.py"
+}

derived/tables/exclusions.jsonl

@@ -0,0 +1,12 @@
+{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 13, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 10, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 2, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 11, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 4, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", "threshold_fraction": 0.01, "total_count": 25}
+{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", "threshold_fraction": 0.01, "total_count": 25}

derived/tables/panels.json

@@ -0,0 +1,28 @@
+{
+  "panels": {
+    "paper_neural_behavior_non_degenerate_v1": {
+      "exclude_by": "exclusions.jsonl reason=paper_degenerate_episode_v1",
+      "table": "rollouts.jsonl",
+      "unit": "checkpoint rollout episode"
+    },
+    "paper_random_opponent_valid_v1": {
+      "include": {
+        "evaluation_protocol": "paper_random_opponent_v1",
+        "status": "ok"
+      },
+      "table": "evaluations.jsonl",
+      "unit": "checkpoint x opponent_mode"
+    },
+    "primary_valid_runs_v1": {
+      "include": {
+        "checkpoint_ok": true,
+        "healthy": true,
+        "is_latest_successful": true,
+        "status": "completed"
+      },
+      "table": "runs.jsonl",
+      "unit": "task x seed latest successful attempt"
+    }
+  },
+  "schema_version": 1
+}

derived/tables/rollouts.jsonl

@@ -0,0 +1,20 @@
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:19.731502+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 1, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:20.616845+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 2, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:21.500858+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 3, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:22.386345+00:00", "degenerate_episode_count": 13, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 4, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:23.279182+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 5, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:24.164102+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 6, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.051469+00:00", "degenerate_episode_count": 10, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 7, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.937613+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 8, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:26.826889+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 9, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:27.714387+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 10, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:28.637667+00:00", "degenerate_episode_count": 2, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 11, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0000"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:29.559684+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 12, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0001"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:30.484944+00:00", "degenerate_episode_count": 11, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 13, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0002"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:31.410494+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 14, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0003"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:32.334168+00:00", "degenerate_episode_count": 4, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 15, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0004"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:33.266584+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 16, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0005"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:34.189914+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 17, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0006"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:35.152778+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 18, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0007"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:36.085812+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 19, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0008"}
+{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:37.008624+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 20, "source_path": "runs/mouse-run-run-2-transformer/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0009"}

derived/tables/runs.jsonl

@@ -0,0 +1,20 @@
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19540.62696071295, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", "schema_version": 1, "seed": 0, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0000", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19646.28115725203, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", "schema_version": 1, "seed": 1, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0001", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19551.709339142893, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", "schema_version": 1, "seed": 2, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0002", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19539.363500279956, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", "schema_version": 1, "seed": 3, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0003", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19559.85987384396, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", "schema_version": 1, "seed": 4, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0004", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19523.735058769933, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", "schema_version": 1, "seed": 5, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0005", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19464.16788898199, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", "schema_version": 1, "seed": 6, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0006", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19529.17995151691, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", "schema_version": 1, "seed": 7, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0007", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19471.327225218993, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", "schema_version": 1, "seed": 8, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0008", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19415.301637928, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", "schema_version": 1, "seed": 9, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0009", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19670.30402108794, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", "schema_version": 1, "seed": 0, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0000", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19568.46814744, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", "schema_version": 1, "seed": 1, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0001", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19568.42091672402, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", "schema_version": 1, "seed": 2, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0002", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19557.178920289967, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", "schema_version": 1, "seed": 3, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0003", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19601.35430799797, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", "schema_version": 1, "seed": 4, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0004", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19533.043117783964, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", "schema_version": 1, "seed": 5, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0005", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19480.819718190003, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", "schema_version": 1, "seed": 6, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0006", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19530.44018116698, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", "schema_version": 1, "seed": 7, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0007", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19499.830016753986, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", "schema_version": 1, "seed": 8, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0008", "update": 20000}
+{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-transformer", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 19455.73302768101, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", "schema_version": 1, "seed": 9, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0009", "update": 20000}

launch_gate.json

@@ -0,0 +1,204 @@
+{
+  "checks": [
+    {
+      "detail": "experiments/paper_marl_2026/SPEC.md",
+      "name": "experiment_spec",
+      "ok": true
+    },
+    {
+      "detail": {
+        "missing": []
+      },
+      "name": "config_schema",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 20000,
+        "expected": 20000
+      },
+      "name": "primary_updates",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 40,
+        "expected": 40
+      },
+      "name": "primary_batch_size",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 100,
+        "expected": 100
+      },
+      "name": "primary_max_steps",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 10,
+        "expected": 10
+      },
+      "name": "primary_seeds",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 0,
+        "expected": 0
+      },
+      "name": "primary_seed_start",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": "both",
+        "expected": "both"
+      },
+      "name": "primary_tasks",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": "paper_text",
+        "expected": "paper_text"
+      },
+      "name": "primary_preset",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": "cuda",
+        "expected": "cuda"
+      },
+      "name": "primary_device",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": true,
+        "expected": true
+      },
+      "name": "primary_finite_guard",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": true,
+        "expected": true
+      },
+      "name": "primary_triton_env_step",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": false,
+        "expected": false
+      },
+      "name": "primary_fused_agent_rollout",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 10,
+        "expected": 10
+      },
+      "name": "primary_subspace_metric_period",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual": 2
+      },
+      "name": "max_attempts",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual_seconds": 1800.0
+      },
+      "name": "observability_period",
+      "ok": true
+    },
+    {
+      "detail": {
+        "actual_hours": 24.0
+      },
+      "name": "attempt_timeout",
+      "ok": true
+    },
+    {
+      "detail": {
+        "cuda_available": true,
+        "device": "cuda"
+      },
+      "name": "device_available",
+      "ok": true
+    },
+    {
+      "detail": {
+        "current_dirty": false,
+        "current_git_sha": "6282c650d281b31d646d537dd3f57c968806a423",
+        "path": "runs/mouse-run-run-2-transformer/triton_equivalence.json",
+        "record_dirty": false,
+        "record_git_sha": "6282c650d281b31d646d537dd3f57c968806a423",
+        "sha_matches": true,
+        "status": "ok"
+      },
+      "name": "triton_equivalence",
+      "ok": true
+    }
+  ],
+  "config": "experiments/paper_marl_2026/configs/arch_transformer.json",
+  "config_sha256": "2bbd0502a075bf0d5c49b248539f3759048c41a1b20e63efbe677e891decf7b1",
+  "created_at": "2026-07-02T14:31:49.723425+00:00",
+  "experiment_spec": "experiments/paper_marl_2026/SPEC.md",
+  "experiment_spec_sha256": "e878bd03a06735c93c8caab4fc769ce63f026a538dfccbc87b588d432e48761e",
+  "provenance": {
+    "command": "scripts/run/launch_gate.py --config experiments/paper_marl_2026/configs/arch_transformer.json --triton-equivalence runs/mouse-run-run-2-transformer/triton_equivalence.json",
+    "created_at": "2026-07-02T14:31:49.723576+00:00",
+    "dependencies": {
+      "huggingface_hub": "1.21.0",
+      "numpy": "2.5.0",
+      "safetensors": "0.8.0",
+      "tensorboard": "2.21.0",
+      "torch": "2.12.1",
+      "triton": "3.7.1"
+    },
+    "executable": "/root/mouse-run-run-perf/.venv/bin/python3",
+    "git": {
+      "branch": "main",
+      "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "dirty": false,
+      "sha": "6282c650d281b31d646d537dd3f57c968806a423",
+      "short_sha": "6282c65",
+      "status_short": ""
+    },
+    "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39",
+    "python": "3.14.5 (main, Jun  2 2026, 22:25:17) [Clang 22.1.3 ]",
+    "schema_version": 1,
+    "torch": {
+      "cuda_available": true,
+      "cuda_device_count": 1,
+      "cuda_devices": [
+        {
+          "index": 0,
+          "major": 8,
+          "minor": 9,
+          "multi_processor_count": 66,
+          "name": "NVIDIA GeForce RTX 4070 Ti SUPER",
+          "total_memory_bytes": 16717578240
+        }
+      ],
+      "cuda_version": "13.0",
+      "cudnn_version": 92000,
+      "mps_available": false,
+      "version": "2.12.1+cu130"
+    }
+  },
+  "schema_version": 1,
+  "status": "ok",
+  "triton_equivalence": "runs/mouse-run-run-2-transformer/triton_equivalence.json"
+}

manifest.json

@@ -0,0 +1,423 @@
+{
+  "config": {
+    "architecture": "transformer",
+    "batch_size": 40,
+    "checkpoint_every": 1000,
+    "checkpoint_every_seconds": 1800.0,
+    "clip_epsilon": 0.2,
+    "concurrency": 10,
+    "cuda_tf32": true,
+    "device": "cuda",
+    "entropy_coef": 0.01,
+    "finite_guard": true,
+    "fused_agent_rollout": false,
+    "grad_clip": 1.0,
+    "hidden_size": 256,
+    "learning_rate": 0.0003,
+    "max_steps": 100,
+    "ppo_epochs": 4,
+    "preset": "paper_text",
+    "recurrent_l2_coef": 0.3,
+    "status_every_seconds": 1800.0,
+    "subspace_metric_period": 10,
+    "triton_env_step": true,
+    "updates": 20000,
+    "value_clip": 10.0,
+    "value_coef": 0.5
+  },
+  "config_sha256": "6e419bd117b1aae2fbbef85bee470cef165112bab56d10a81e5dfbdc5bb6f9c5",
+  "created_at": "2026-07-02T14:31:50.715150+00:00",
+  "experiment": "mouse-run-run-2-transformer",
+  "experiment_spec": "experiments/paper_marl_2026/SPEC.md",
+  "experiment_spec_sha256": "e878bd03a06735c93c8caab4fc769ce63f026a538dfccbc87b588d432e48761e",
+  "initial_attempts": [
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01",
+      "seed": 0,
+      "task": "social",
+      "unit_id": "social/seed_0000"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01",
+      "seed": 0,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0000"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01",
+      "seed": 1,
+      "task": "social",
+      "unit_id": "social/seed_0001"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01",
+      "seed": 1,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0001"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01",
+      "seed": 2,
+      "task": "social",
+      "unit_id": "social/seed_0002"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01",
+      "seed": 2,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0002"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01",
+      "seed": 3,
+      "task": "social",
+      "unit_id": "social/seed_0003"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01",
+      "seed": 3,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0003"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01",
+      "seed": 4,
+      "task": "social",
+      "unit_id": "social/seed_0004"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01",
+      "seed": 4,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0004"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01",
+      "seed": 5,
+      "task": "social",
+      "unit_id": "social/seed_0005"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01",
+      "seed": 5,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0005"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01",
+      "seed": 6,
+      "task": "social",
+      "unit_id": "social/seed_0006"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01",
+      "seed": 6,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0006"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01",
+      "seed": 7,
+      "task": "social",
+      "unit_id": "social/seed_0007"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01",
+      "seed": 7,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0007"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01",
+      "seed": 8,
+      "task": "social",
+      "unit_id": "social/seed_0008"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01",
+      "seed": 8,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0008"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01",
+      "seed": 9,
+      "task": "social",
+      "unit_id": "social/seed_0009"
+    },
+    {
+      "attempt": 1,
+      "attempt_id": "attempt_01",
+      "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors",
+      "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log",
+      "resume_from": null,
+      "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01",
+      "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01",
+      "seed": 9,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0009"
+    }
+  ],
+  "max_attempts": 2,
+  "provenance": {
+    "command": "scripts/run_local_experiment.py --config experiments/paper_marl_2026/configs/arch_transformer.json",
+    "created_at": "2026-07-02T14:31:50.715276+00:00",
+    "dependencies": {
+      "huggingface_hub": "1.21.0",
+      "numpy": "2.5.0",
+      "safetensors": "0.8.0",
+      "tensorboard": "2.21.0",
+      "torch": "2.12.1",
+      "triton": "3.7.1"
+    },
+    "executable": "/root/mouse-run-run-perf/.venv/bin/python3",
+    "git": {
+      "branch": "main",
+      "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+      "dirty": false,
+      "sha": "6282c650d281b31d646d537dd3f57c968806a423",
+      "short_sha": "6282c65",
+      "status_short": ""
+    },
+    "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39",
+    "python": "3.14.5 (main, Jun  2 2026, 22:25:17) [Clang 22.1.3 ]",
+    "schema_version": 1,
+    "torch": {
+      "cuda_available": true,
+      "cuda_device_count": 1,
+      "cuda_devices": [
+        {
+          "index": 0,
+          "major": 8,
+          "minor": 9,
+          "multi_processor_count": 66,
+          "name": "NVIDIA GeForce RTX 4070 Ti SUPER",
+          "total_memory_bytes": 16717578240
+        }
+      ],
+      "cuda_version": "13.0",
+      "cudnn_version": 92000,
+      "mps_available": false,
+      "version": "2.12.1+cu130"
+    }
+  },
+  "schema_version": 2,
+  "total_environment_steps": 1600000000,
+  "total_units": 20,
+  "units": [
+    {
+      "seed": 0,
+      "task": "social",
+      "unit_id": "social/seed_0000"
+    },
+    {
+      "seed": 0,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0000"
+    },
+    {
+      "seed": 1,
+      "task": "social",
+      "unit_id": "social/seed_0001"
+    },
+    {
+      "seed": 1,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0001"
+    },
+    {
+      "seed": 2,
+      "task": "social",
+      "unit_id": "social/seed_0002"
+    },
+    {
+      "seed": 2,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0002"
+    },
+    {
+      "seed": 3,
+      "task": "social",
+      "unit_id": "social/seed_0003"
+    },
+    {
+      "seed": 3,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0003"
+    },
+    {
+      "seed": 4,
+      "task": "social",
+      "unit_id": "social/seed_0004"
+    },
+    {
+      "seed": 4,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0004"
+    },
+    {
+      "seed": 5,
+      "task": "social",
+      "unit_id": "social/seed_0005"
+    },
+    {
+      "seed": 5,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0005"
+    },
+    {
+      "seed": 6,
+      "task": "social",
+      "unit_id": "social/seed_0006"
+    },
+    {
+      "seed": 6,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0006"
+    },
+    {
+      "seed": 7,
+      "task": "social",
+      "unit_id": "social/seed_0007"
+    },
+    {
+      "seed": 7,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0007"
+    },
+    {
+      "seed": 8,
+      "task": "social",
+      "unit_id": "social/seed_0008"
+    },
+    {
+      "seed": 8,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0008"
+    },
+    {
+      "seed": 9,
+      "task": "social",
+      "unit_id": "social/seed_0009"
+    },
+    {
+      "seed": 9,
+      "task": "non_social",
+      "unit_id": "non_social/seed_0009"
+    }
+  ]
+}

non_social/seed_0000/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bebcbee7c71f93593b985b9403dddac3f4b9d301088feea80adec365e5c839d3
+size 29760368

non_social/seed_0000/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01",
+  "seed": 0,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0000/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0007858585283328365, "elapsed_seconds": 2.8290907019982114, "env_steps_per_second": 1413.8818515697517, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 15.716384708128398, "eta_seconds": 56578.98494926223, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.003984742708620814, "chaser_grad_norm": 6.097266681492329, "chaser_new_fields": 34.67499923706055, "chaser_partner_vision": 0.0, "chaser_return": -3.130000352859497, "chaser_subspace_norm": 0.03538785129785538, "collisions_per_episode": 2.1500000953674316, "entropy": 2.757209300994873, "explorer_grad_norm": 6.22088485956192, "explorer_new_fields": 34.04999923706055, "explorer_partner_vision": 0.0, "explorer_return": 0.512499988079071, "explorer_subspace_norm": 0.0416051484644413, "final_distance": 5.596360206604004, "max_param_abs": 1.001202940940857, "policy_loss": -0.015378041312104074, "value_loss": 7.977049231529236}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.904332+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.9043489, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.13819628580276105, "elapsed_seconds": 497.50662888993975, "env_steps_per_second": 4020.0469377915515, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.389655146307681, "eta_seconds": 19402.75852670765, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005213568198339402, "chaser_grad_norm": 0.6902080327272415, "chaser_new_fields": 85.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.1100006103515625, "chaser_subspace_norm": 12.22103214263916, "collisions_per_episode": 1.399999976158142, "entropy": 0.10881881602108479, "explorer_grad_norm": 4.12031877040863, "explorer_new_fields": 89.875, "explorer_partner_vision": 0.0, "explorer_return": 84.4375, "explorer_subspace_norm": 15.985239028930664, "final_distance": 3.488882541656494, "max_param_abs": 1.2841757535934448, "policy_loss": -0.003982506634656602, "value_loss": 0.554658830165863}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:09.581869+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003209.5818808, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.27406795207886414, "elapsed_seconds": 986.6446274839109, "env_steps_per_second": 4054.144611520958, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.207291089498418, "eta_seconds": 18746.247922194307, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0009871974141866247, "chaser_grad_norm": 0.08980551920831203, "chaser_new_fields": 88.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.755000591278076, "chaser_subspace_norm": 17.67871856689453, "collisions_per_episode": 0.7750000357627869, "entropy": 0.03490936104208231, "explorer_grad_norm": 10.575141370296478, "explorer_new_fields": 93.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 89.3499984741211, "explorer_subspace_norm": 29.749900817871094, "final_distance": 3.751117706298828, "max_param_abs": 1.4041469097137451, "policy_loss": -0.002187443812843526, "value_loss": 0.4212941825389862}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:18.719867+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003698.719881, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.4078779447350017, "elapsed_seconds": 1468.3606010460062, "env_steps_per_second": 4086.1897245988625, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.030494651731687, "eta_seconds": 18109.780746234075, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0010331508172161108, "chaser_grad_norm": 0.11727817915380001, "chaser_new_fields": 90.625, "chaser_partner_vision": 0.0, "chaser_return": 8.120000839233398, "chaser_subspace_norm": 22.310590744018555, "collisions_per_episode": 0.22500000894069672, "entropy": 0.01846518414095044, "explorer_grad_norm": 3.1013119220733643, "explorer_new_fields": 94.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.26250457763672, "explorer_subspace_norm": 40.93424606323242, "final_distance": 3.4480628967285156, "max_param_abs": 1.505894422531128, "policy_loss": -0.0012613329393786898, "value_loss": 0.07974300533533096}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:20.435839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004180.43585, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5416774474972044, "elapsed_seconds": 1950.0388109899359, "env_steps_per_second": 4102.482450561487, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8750970274748395, "eta_seconds": 17550.349298909423, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0010745732343774295, "chaser_grad_norm": 0.32581115886569023, "chaser_new_fields": 87.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.460000514984131, "chaser_subspace_norm": 26.053730010986328, "collisions_per_episode": 1.524999976158142, "entropy": 0.04285650607198477, "explorer_grad_norm": 2.352870851755142, "explorer_new_fields": 92.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 88.11250305175781, "explorer_subspace_norm": 48.09144973754883, "final_distance": 3.754744052886963, "max_param_abs": 1.5702295303344727, "policy_loss": -0.0025823999991416713, "value_loss": 0.5012058317661285}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:22.114052+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004662.1140668, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6754466422619427, "elapsed_seconds": 2431.607912142994, "env_steps_per_second": 4112.50512472092, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.7281264958336, "eta_seconds": 17021.255385000957, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0009930499389909642, "chaser_grad_norm": 0.04422433767467737, "chaser_new_fields": 89.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.875000953674316, "chaser_subspace_norm": 27.880756378173828, "collisions_per_episode": 0.7750000357627869, "entropy": 0.016509887762367725, "explorer_grad_norm": 1.9961952567100525, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 57.431854248046875, "final_distance": 4.026604652404785, "max_param_abs": 1.6253595352172852, "policy_loss": -0.0012972251977743099, "value_loss": 0.20397597178816795}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:23.683150+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005143.6831622, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8092757643613732, "elapsed_seconds": 2913.3927517009433, "env_steps_per_second": 4118.909128538872, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.585895998047781, "eta_seconds": 16509.225592972012, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0027243835385055704, "chaser_grad_norm": 0.4044172838330269, "chaser_new_fields": 90.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.135001182556152, "chaser_subspace_norm": 29.07428550720215, "collisions_per_episode": 0.3499999940395355, "entropy": 0.013051415095105767, "explorer_grad_norm": 0.7505706623196602, "explorer_new_fields": 94.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 68.80549621582031, "final_distance": 3.6912147998809814, "max_param_abs": 1.6739788055419922, "policy_loss": -0.003254898154175123, "value_loss": 0.12792102247476578}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:25.467991+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005625.4680054, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9452245050919332, "elapsed_seconds": 3402.80821833096, "env_steps_per_second": 4114.248909057486, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.4560583811476855, "eta_seconds": 16041.810172131667, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0001970463801899136, "chaser_grad_norm": 0.0663881553336978, "chaser_new_fields": 90.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.095001220703125, "chaser_subspace_norm": 34.47535705566406, "collisions_per_episode": 0.699999988079071, "entropy": 0.009223120054230094, "explorer_grad_norm": 0.439394012093544, "explorer_new_fields": 94.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 71.96785736083984, "final_distance": 4.070150852203369, "max_param_abs": 1.7069555521011353, "policy_loss": -0.0004109265974836518, "value_loss": 0.12886791303753853}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:34.883457+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006114.8834677, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0821224993897098, "elapsed_seconds": 3895.640997802955, "env_steps_per_second": 4107.1546400255065, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.3284899975588385, "eta_seconds": 15582.563991211819, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.002061240742013748, "chaser_grad_norm": 0.14911585487425327, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 36.80433654785156, "collisions_per_episode": 0.800000011920929, "entropy": 0.007869288441725075, "explorer_grad_norm": 0.9927732795476913, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 77.96184539794922, "final_distance": 3.4226481914520264, "max_param_abs": 1.741771936416626, "policy_loss": -0.003020491153939986, "value_loss": 0.147367924451828}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:47.716237+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006607.716251, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.2179189361097007, "elapsed_seconds": 4384.508169994922, "env_steps_per_second": 4105.363544121494, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.195054113266747, "eta_seconds": 15102.194807760288, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0016153937390637196, "chaser_grad_norm": 0.06304028397426009, "chaser_new_fields": 93.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 38.633880615234375, "collisions_per_episode": 0.800000011920929, "entropy": 0.00674564519431442, "explorer_grad_norm": 0.7198946475982666, "explorer_new_fields": 94.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 83.59247589111328, "final_distance": 3.769803762435913, "max_param_abs": 1.7817188501358032, "policy_loss": -0.0006622512721223828, "value_loss": 0.11152646876871586}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:44:56.583410+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007096.5834255, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.357942673002221, "elapsed_seconds": 4888.593622807995, "env_steps_per_second": 4091.156177655866, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.073828019006663, "eta_seconds": 14665.780868423986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0002423651793477255, "chaser_grad_norm": 0.038826094940304756, "chaser_new_fields": 92.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.480000495910645, "chaser_subspace_norm": 37.68648910522461, "collisions_per_episode": 0.6000000238418579, "entropy": 0.00865237764082849, "explorer_grad_norm": 0.6159049421548843, "explorer_new_fields": 94.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 86.39503479003906, "final_distance": 3.6380271911621094, "max_param_abs": 1.810005784034729, "policy_loss": -0.0006053061515558511, "value_loss": 0.08581102639436722}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:20.668862+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007600.6688757, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.496551224394158, "elapsed_seconds": 5387.584407818969, "env_steps_per_second": 4083.4627051172565, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9454532279482346, "eta_seconds": 14203.631620613645, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0007384293112006057, "chaser_grad_norm": 0.12568512931466103, "chaser_new_fields": 92.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.55000114440918, "chaser_subspace_norm": 41.24983215332031, "collisions_per_episode": 0.45000001788139343, "entropy": 0.008910412434488535, "explorer_grad_norm": 0.9198387712240219, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.2750015258789, "explorer_subspace_norm": 90.77873992919922, "final_distance": 3.517383575439453, "max_param_abs": 1.8441685438156128, "policy_loss": -0.0011138532813639301, "value_loss": 0.06246422231197357}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:39.659647+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008099.6596582, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.6330133805391638, "elapsed_seconds": 5878.8481699409895, "env_steps_per_second": 4082.4323585551806, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.8103645545913816, "eta_seconds": 13717.312396528974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0002877887917747124, "chaser_grad_norm": 0.0709311356768012, "chaser_new_fields": 92.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.465001106262207, "chaser_subspace_norm": 39.979000091552734, "collisions_per_episode": 0.75, "entropy": 0.006909179617650807, "explorer_grad_norm": 0.5209658145904541, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.67500305175781, "explorer_subspace_norm": 97.17213439941406, "final_distance": 3.5823585987091064, "max_param_abs": 1.865701675415039, "policy_loss": -0.0007798157297145281, "value_loss": 0.11099696718156338}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:50.923411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008590.9234254, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.767157393639144, "elapsed_seconds": 6361.766617100919, "env_steps_per_second": 4086.915092123939, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6702499714043766, "eta_seconds": 13212.899897055755, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0007893479407814347, "chaser_grad_norm": 0.04300319030880928, "chaser_new_fields": 92.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.510001182556152, "chaser_subspace_norm": 45.24545669555664, "collisions_per_episode": 0.625, "entropy": 0.005201557534746826, "explorer_grad_norm": 0.36915695667266846, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.01250457763672, "explorer_subspace_norm": 100.9892349243164, "final_distance": 3.551710605621338, "max_param_abs": 1.8893122673034668, "policy_loss": -0.00030515577702328045, "value_loss": 0.10670126788318157}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:53.841855+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009073.8418665, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.9058069222163694, "elapsed_seconds": 6860.90491997893, "env_steps_per_second": 4081.094305572448, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5393557126875432, "eta_seconds": 12741.680565675155, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005970458028607961, "chaser_grad_norm": 0.08210457861423492, "chaser_new_fields": 92.625, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 44.65858459472656, "collisions_per_episode": 0.5, "entropy": 0.006243823445402086, "explorer_grad_norm": 0.2898806743323803, "explorer_new_fields": 95.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 101.96656036376953, "final_distance": 3.8500595092773438, "max_param_abs": 1.9082915782928467, "policy_loss": -0.0005075841061774611, "value_loss": 0.06872157379984856}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:26:12.980159+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009572.9801733, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.0401156900350017, "elapsed_seconds": 7344.416484126006, "env_steps_per_second": 4084.735671627701, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.400192816725003, "eta_seconds": 12240.694140210011, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0012023549742483652, "chaser_grad_norm": 0.18835537135601044, "chaser_new_fields": 92.875, "chaser_partner_vision": 0.0, "chaser_return": 8.51500129699707, "chaser_subspace_norm": 47.02663040161133, "collisions_per_episode": 0.875, "entropy": 0.004316961974836886, "explorer_grad_norm": 0.742193266749382, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 106.36807250976562, "final_distance": 3.8010079860687256, "max_param_abs": 1.9282002449035645, "policy_loss": -0.0011232557701852386, "value_loss": 0.10672311298549175}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:16.491724+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010056.491738, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.178467757531649, "elapsed_seconds": 7842.483927113935, "env_steps_per_second": 4080.33988942788, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.267701636297473, "eta_seconds": 11763.725890670903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -1.6572943833947917e-05, "chaser_grad_norm": 0.03920569969341159, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 47.99747085571289, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004179352545179427, "explorer_grad_norm": 0.42505457252264023, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.5250015258789, "explorer_subspace_norm": 108.12300872802734, "final_distance": 3.975580930709839, "max_param_abs": 1.9446109533309937, "policy_loss": -8.778000239839656e-05, "value_loss": 0.11882751807570457}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:34.559169+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010554.5591846, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.3151148458847053, "elapsed_seconds": 8334.413445184939, "env_steps_per_second": 4079.4712457710993, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1322142032557774, "eta_seconds": 11275.971131720798, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00039934206656147553, "chaser_grad_norm": 0.054306747391819954, "chaser_new_fields": 93.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 44.896263122558594, "collisions_per_episode": 0.675000011920929, "entropy": 0.006878925370983779, "explorer_grad_norm": 0.43170301616191864, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 110.83919525146484, "final_distance": 4.0531816482543945, "max_param_abs": 1.9534330368041992, "policy_loss": -0.0007233839633471462, "value_loss": 0.06848301365971565}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:46.488683+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011046.4886954, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.449597211649169, "elapsed_seconds": 8818.54996193701, "env_steps_per_second": 4082.3037977201116, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.993952147571207, "eta_seconds": 10778.227731256346, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00018719192692023823, "chaser_grad_norm": 0.11254973709583282, "chaser_new_fields": 93.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.61500072479248, "chaser_subspace_norm": 48.95949935913086, "collisions_per_episode": 0.875, "entropy": 0.0051835766062140465, "explorer_grad_norm": 0.39654700085520744, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.1624984741211, "explorer_subspace_norm": 114.00794982910156, "final_distance": 4.262425899505615, "max_param_abs": 1.9768918752670288, "policy_loss": -0.0004432122928577087, "value_loss": 0.06671177595853806}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:50.625202+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011530.6252155, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.5876557905505373, "elapsed_seconds": 9315.560845981934, "env_steps_per_second": 4079.196156653357, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.860040610608489, "eta_seconds": 10296.14619819056, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005332294335725907, "chaser_grad_norm": 0.0458578048273921, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.59000015258789, "chaser_subspace_norm": 50.698280334472656, "collisions_per_episode": 0.550000011920929, "entropy": 0.00311667169444263, "explorer_grad_norm": 0.554467223584652, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 115.92314910888672, "final_distance": 4.192587375640869, "max_param_abs": 1.9959673881530762, "policy_loss": -0.00047228195774096093, "value_loss": 0.05310062225908041}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:07:07.636087+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012027.6361022, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.7234276721347124, "elapsed_seconds": 9804.339619684964, "env_steps_per_second": 4079.826031290141, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7234276721347124, "eta_seconds": 9804.339619684964, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005473158898638975, "chaser_grad_norm": 0.12831407506018877, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 52.909427642822266, "collisions_per_episode": 0.699999988079071, "entropy": 0.0029723012703470886, "explorer_grad_norm": 0.2813072130084038, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 116.91361236572266, "final_distance": 4.026191234588623, "max_param_abs": 2.006232500076294, "policy_loss": -0.000196251879156506, "value_loss": 0.07602320238947868}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:15:16.414861+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012516.4148755, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.8574518892694485, "elapsed_seconds": 10286.826801370014, "env_steps_per_second": 4082.8917226842373, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.58531361410093, "eta_seconds": 9307.129010763347, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0009713084957361506, "chaser_grad_norm": 0.027999839279800653, "chaser_new_fields": 93.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 54.712059020996094, "collisions_per_episode": 0.75, "entropy": 0.005503859603777528, "explorer_grad_norm": 0.2878112308681011, "explorer_new_fields": 94.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 120.97154235839844, "final_distance": 3.91410756111145, "max_param_abs": 2.0216140747070312, "policy_loss": -0.0005653448201483613, "value_loss": 0.1004257034510374}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:18.902040+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012998.902052, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.9921132260405447, "elapsed_seconds": 10771.60761374596, "env_steps_per_second": 4084.812738987106, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4480926394877183, "eta_seconds": 8813.133502155786, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 1.541443885122451e-05, "chaser_grad_norm": 0.029784617945551872, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 53.42091751098633, "collisions_per_episode": 0.625, "entropy": 0.0029498518561013043, "explorer_grad_norm": 0.29834797978401184, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 121.83349609375, "final_distance": 3.73388671875, "max_param_abs": 2.040980339050293, "policy_loss": -9.137296861494448e-05, "value_loss": 0.066783607006073}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:23.682854+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013483.6828663, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.1278307616852623, "elapsed_seconds": 11260.190742066945, "env_steps_per_second": 4085.188346601324, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3118749108108463, "eta_seconds": 8322.749678919046, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00011620879817808571, "chaser_grad_norm": 0.09901169408112764, "chaser_new_fields": 93.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 54.40679168701172, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004110373207367957, "explorer_grad_norm": 0.2696717716753483, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 123.82683563232422, "final_distance": 3.829066514968872, "max_param_abs": 2.0539538860321045, "policy_loss": -0.000206822176653354, "value_loss": 0.07687818817794323}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:32.265981+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013972.2659926, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.262044489343049, "elapsed_seconds": 11743.360161634977, "env_steps_per_second": 4087.41615170876, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1746963262286996, "eta_seconds": 7828.906774423318, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00032187812307515884, "chaser_grad_norm": 0.04246761417016387, "chaser_new_fields": 93.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 56.21242141723633, "collisions_per_episode": 0.75, "entropy": 0.003060928371269256, "explorer_grad_norm": 0.5045444145798683, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 125.90714263916016, "final_distance": 3.640484571456909, "max_param_abs": 2.057662010192871, "policy_loss": -0.00028770001560118885, "value_loss": 0.07355882227420807}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:35.435403+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014455.4354193, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.396113327720539, "elapsed_seconds": 12226.00797979394, "env_steps_per_second": 4089.6423495416952, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0376679966323232, "eta_seconds": 7335.604787876364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006283227894127719, "chaser_grad_norm": 0.1221990492194891, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.644999504089355, "chaser_subspace_norm": 56.58079528808594, "collisions_per_episode": 0.7750000357627869, "entropy": 0.005023272125981748, "explorer_grad_norm": 0.3120904453098774, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 127.02792358398438, "final_distance": 4.106462001800537, "max_param_abs": 2.0744540691375732, "policy_loss": -0.0007666714633818295, "value_loss": 0.0993267111480236}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:38.083219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014938.08323, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5320635015466704, "elapsed_seconds": 12715.428605568013, "env_steps_per_second": 4089.5200321623056, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9018803469866685, "eta_seconds": 6846.7692491520065, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00033326833698955893, "chaser_grad_norm": 0.04215310048311949, "chaser_new_fields": 93.625, "chaser_partner_vision": 0.0, "chaser_return": 8.710000038146973, "chaser_subspace_norm": 56.401405334472656, "collisions_per_episode": 0.625, "entropy": 0.005254154675640166, "explorer_grad_norm": 0.2833205871284008, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 127.3544921875, "final_distance": 4.4870429039001465, "max_param_abs": 2.089384078979492, "policy_loss": -0.0005055732870147267, "value_loss": 0.09795444644987583}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:47.503846+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015427.5038593, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.668322203079426, "elapsed_seconds": 13205.959931085934, "env_steps_per_second": 4089.0628384300685, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.766229208890094, "eta_seconds": 6358.425152004338, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00034709783561170937, "chaser_grad_norm": 0.0546520808711648, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 56.26631546020508, "collisions_per_episode": 0.625, "entropy": 0.004097803262993693, "explorer_grad_norm": 0.46600425988435745, "explorer_new_fields": 94.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 132.22207641601562, "final_distance": 3.958134412765503, "max_param_abs": 2.109316825866699, "policy_loss": -0.000491810953768379, "value_loss": 0.06270698457956314}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:58.035172+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015918.0351858, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.805968612701658, "elapsed_seconds": 13701.487005725969, "env_steps_per_second": 4087.1476195683813, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6311294054435677, "eta_seconds": 5872.065859596843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.000709932864020546, "chaser_grad_norm": 0.060600987169891596, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 59.46336364746094, "collisions_per_episode": 0.574999988079071, "entropy": 0.004734171321615577, "explorer_grad_norm": 0.20277123898267746, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 138.1077117919922, "final_distance": 3.830503225326538, "max_param_abs": 2.1345181465148926, "policy_loss": -0.0005144484461050425, "value_loss": 0.0775738526135683}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:13.562247+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016413.562261, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.940078420972762, "elapsed_seconds": 14184.282315501943, "env_steps_per_second": 4089.033107907902, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4945125045069099, "eta_seconds": 5380.245016224875, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003340224873058517, "chaser_grad_norm": 0.07462147623300552, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 59.58805465698242, "collisions_per_episode": 0.574999988079071, "entropy": 0.005957852001301944, "explorer_grad_norm": 0.32502011209726334, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 137.80844116210938, "final_distance": 3.650078535079956, "max_param_abs": 2.134597063064575, "policy_loss": -0.0004052768719855493, "value_loss": 0.08263413421809673}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:16.357554+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016896.357567, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.074239472052755, "elapsed_seconds": 14667.262099389918, "env_steps_per_second": 4090.7430162099367, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3580798240175849, "eta_seconds": 4889.087366463305, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.000514724589585569, "chaser_grad_norm": 0.053545817732810974, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 63.33787536621094, "collisions_per_episode": 0.550000011920929, "entropy": 0.004608422867022455, "explorer_grad_norm": 0.30683189257979393, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.32500457763672, "explorer_subspace_norm": 137.96897888183594, "final_distance": 3.8315236568450928, "max_param_abs": 2.139131546020508, "policy_loss": -0.0006109134045004794, "value_loss": 0.056757569313049316}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:19.337340+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017379.337354, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.210373383232218, "elapsed_seconds": 15157.344179635984, "env_steps_per_second": 4090.4263481261783, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2223664660996763, "eta_seconds": 4400.519277958834, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004188151698392151, "chaser_grad_norm": 0.03207044675946236, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.90999984741211, "chaser_subspace_norm": 61.26824188232422, "collisions_per_episode": 0.800000011920929, "entropy": 0.0030996299465186894, "explorer_grad_norm": 0.38353733718395233, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 139.2194061279297, "final_distance": 3.9234397411346436, "max_param_abs": 2.1418116092681885, "policy_loss": -0.00018357145809488884, "value_loss": 0.05587191041558981}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:29.419418+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017869.4194288, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.345209612597763, "elapsed_seconds": 15642.754605351947, "env_steps_per_second": 4091.3510193468933, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0863024031494408, "eta_seconds": 3910.688651337987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00031747748237398454, "chaser_grad_norm": 0.05185403674840927, "chaser_new_fields": 94.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.9399995803833, "chaser_subspace_norm": 64.73976135253906, "collisions_per_episode": 0.7750000357627869, "entropy": 0.004601132357493043, "explorer_grad_norm": 0.21299517154693604, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 139.50474548339844, "final_distance": 3.724943161010742, "max_param_abs": 2.1471619606018066, "policy_loss": -0.000248220444746039, "value_loss": 0.07776057720184326}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:34.829844+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018354.8298604, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.479314453306369, "elapsed_seconds": 16125.532031902927, "env_steps_per_second": 4092.8882141330214, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9501576113074115, "eta_seconds": 3420.5674007066814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00019849943035726847, "chaser_grad_norm": 0.04664633749052882, "chaser_new_fields": 95.0, "chaser_partner_vision": 0.0, "chaser_return": 8.975000381469727, "chaser_subspace_norm": 68.20988464355469, "collisions_per_episode": 0.7250000238418579, "entropy": 0.002828369732014835, "explorer_grad_norm": 0.5940320566296577, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 143.01023864746094, "final_distance": 3.92901611328125, "max_param_abs": 2.1550333499908447, "policy_loss": -0.0003368458753683079, "value_loss": 0.0917948242276907}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:37.607270+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018837.6072822, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.61318730738415, "elapsed_seconds": 16607.47430658294, "env_steps_per_second": 4094.541936039358, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8140918777736734, "eta_seconds": 2930.7307599852243, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00021912654095368467, "chaser_grad_norm": 0.05596489645540714, "chaser_new_fields": 95.0, "chaser_partner_vision": 0.0, "chaser_return": 8.960000038146973, "chaser_subspace_norm": 67.79967498779297, "collisions_per_episode": 0.45000001788139343, "entropy": 0.003220201935619116, "explorer_grad_norm": 0.2399616837501526, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.25, "explorer_subspace_norm": 141.00111389160156, "final_distance": 3.815213918685913, "max_param_abs": 2.1613283157348633, "policy_loss": -0.00019624019481145893, "value_loss": 0.04436328634619713}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:39.549546+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019319.5495608, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.74989137156611, "elapsed_seconds": 17099.608937637997, "env_steps_per_second": 4093.660869981816, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6785559102237301, "eta_seconds": 2442.801276805428, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 5.842784732817963e-05, "chaser_grad_norm": 0.02148897247388959, "chaser_new_fields": 95.125, "chaser_partner_vision": 0.0, "chaser_return": 8.970000267028809, "chaser_subspace_norm": 69.08584594726562, "collisions_per_episode": 0.75, "entropy": 0.0033080510911531746, "explorer_grad_norm": 0.2629040814936161, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 141.89207458496094, "final_distance": 3.8556153774261475, "max_param_abs": 2.177971124649048, "policy_loss": -7.839060096737072e-05, "value_loss": 0.08136884681880474}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:51.684176+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019811.684187, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.88537475555081, "elapsed_seconds": 17587.349119982915, "env_steps_per_second": 4093.8517515520807, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428194172834233, "eta_seconds": 1954.1499022203238, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005076670059834323, "chaser_grad_norm": 0.03680429421365261, "chaser_new_fields": 95.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.994999885559082, "chaser_subspace_norm": 62.626258850097656, "collisions_per_episode": 0.574999988079071, "entropy": 0.003358416724950075, "explorer_grad_norm": 0.37850791588425636, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 145.23422241210938, "final_distance": 4.000051975250244, "max_param_abs": 2.179173469543457, "policy_loss": -0.0003006897167652234, "value_loss": 0.05968277156352997}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:24:59.424359+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020299.4243724, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.020150505876939, "elapsed_seconds": 18072.541821156978, "env_steps_per_second": 4094.6094208713043, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4070392302062382, "eta_seconds": 1465.3412287424576, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 1.8009533259876614e-05, "chaser_grad_norm": 0.03915706975385547, "chaser_new_fields": 94.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.894999504089355, "chaser_subspace_norm": 67.8890380859375, "collisions_per_episode": 0.675000011920929, "entropy": 0.001400032255332917, "explorer_grad_norm": 0.17385003715753555, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 142.23980712890625, "final_distance": 3.8947174549102783, "max_param_abs": 2.1906604766845703, "policy_loss": -3.2025814260094876e-05, "value_loss": 0.06482306495308876}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:04.617059+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020784.6170716, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.156070570738876, "elapsed_seconds": 18561.854054659954, "env_steps_per_second": 4094.4185735002156, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2713721353020461, "eta_seconds": 976.939687087366, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00017792411029660116, "chaser_grad_norm": 0.03782106749713421, "chaser_new_fields": 95.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.970000267028809, "chaser_subspace_norm": 69.94136810302734, "collisions_per_episode": 0.4749999940395355, "entropy": 0.002133544418029487, "explorer_grad_norm": 0.33831240609288216, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 150.135009765625, "final_distance": 3.7090566158294678, "max_param_abs": 2.208369493484497, "policy_loss": -8.993363421172518e-05, "value_loss": 0.04737705644220114}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:13.929295+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021273.9293084, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.290843675516112, "elapsed_seconds": 19047.037231858005, "env_steps_per_second": 4095.1250869367486, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.135662658346567, "eta_seconds": 488.3855700476411, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0016687422878480795, "chaser_grad_norm": 0.2595461467280984, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.909998893737793, "chaser_subspace_norm": 70.38494110107422, "collisions_per_episode": 0.9750000238418579, "entropy": 0.003381687856744975, "explorer_grad_norm": 0.2522974908351898, "explorer_new_fields": 94.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.86250305175781, "explorer_subspace_norm": 155.3250732421875, "final_distance": 3.5724456310272217, "max_param_abs": 2.1981301307678223, "policy_loss": -0.0008378539641853422, "value_loss": 0.11335896141827106}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:19.112472+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021759.1124866, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.42705643406639, "elapsed_seconds": 19537.403162639006, "env_steps_per_second": 4094.709994672293, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:29.478401+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022249.4784148, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.427951933531375, "elapsed_seconds": 19540.62696071295, "env_steps_per_second": 4094.034452468825, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:32.702201+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022252.702217, "update": 20000}

non_social/seed_0000/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.427951933531375,
+  "elapsed_seconds": 19540.62696071295,
+  "env_steps_per_second": 4094.034452468825,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 0.0003983886290532812,
+    "chaser_grad_norm": 0.02647677343338728,
+    "chaser_new_fields": 94.9000015258789,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.955000877380371,
+    "chaser_subspace_norm": 75.5232925415039,
+    "collisions_per_episode": 0.550000011920929,
+    "entropy": 0.0030606171931140125,
+    "explorer_grad_norm": 0.17282858304679394,
+    "explorer_new_fields": 95.42500305175781,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 92.95000457763672,
+    "explorer_subspace_norm": 155.51156616210938,
+    "final_distance": 3.808248281478882,
+    "max_param_abs": 2.2150280475616455,
+    "policy_loss": -0.00044627454370527175,
+    "value_loss": 0.06406432762742043
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-02T19:57:32.702201+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783022252.702217,
+  "update": 20000
+}

non_social/seed_0001/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97e55e51803fcb1f12344ef9426ee05617a44dd6420294c15ba2d71524feac2c
+size 29760368

non_social/seed_0001/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01",
+  "seed": 1,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0001/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0007073184386050949, "elapsed_seconds": 2.546346378978342, "env_steps_per_second": 1570.8781935648915, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.145661453663294, "eta_seconds": 50924.38123318786, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006091276896187381, "chaser_grad_norm": 4.504279673099518, "chaser_new_fields": 34.10000228881836, "chaser_partner_vision": 0.0, "chaser_return": -3.2150001525878906, "chaser_subspace_norm": 0.03877149149775505, "collisions_per_episode": 1.225000023841858, "entropy": 2.748730421066284, "explorer_grad_norm": 9.367615342140198, "explorer_new_fields": 35.10000228881836, "explorer_partner_vision": 0.0, "explorer_return": 2.4625000953674316, "explorer_subspace_norm": 0.030939258635044098, "final_distance": 5.002046585083008, "max_param_abs": 1.0012006759643555, "policy_loss": -0.013167476201429551, "value_loss": 7.158067882061005}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.483055+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.483091, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.13894482685805998, "elapsed_seconds": 500.2013766890159, "env_steps_per_second": 3998.389635067789, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.4188482474643385, "eta_seconds": 19507.85369087162, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010831825758677027, "chaser_grad_norm": 0.3167974688112736, "chaser_new_fields": 90.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.170001029968262, "chaser_subspace_norm": 12.111225128173828, "collisions_per_episode": 0.574999988079071, "entropy": 0.10503689385950565, "explorer_grad_norm": 10.194164991378784, "explorer_new_fields": 88.5, "explorer_partner_vision": 0.0, "explorer_return": 82.5999984741211, "explorer_subspace_norm": 14.751652717590332, "final_distance": 3.2616829872131348, "max_param_abs": 1.3052014112472534, "policy_loss": -0.001630453222198014, "value_loss": 0.521367721259594}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:12.138077+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003212.1380937, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.27813606615220327, "elapsed_seconds": 1001.2898381479317, "env_steps_per_second": 3994.8472935655973, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.284585256891862, "eta_seconds": 19024.506924810703, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.028383875501535982, "chaser_grad_norm": 1.810861423611641, "chaser_new_fields": 82.0, "chaser_partner_vision": 0.0, "chaser_return": 6.39500093460083, "chaser_subspace_norm": 17.26828384399414, "collisions_per_episode": 1.524999976158142, "entropy": 0.02800812106579542, "explorer_grad_norm": 2.9772969484329224, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.17500305175781, "explorer_subspace_norm": 28.539579391479492, "final_distance": 2.783372640609741, "max_param_abs": 1.520729899406433, "policy_loss": 2.5617609253458795e-05, "value_loss": 0.20261814072728157}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:33.226534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003713.2265503, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.4171141794858138, "elapsed_seconds": 1501.6110461489297, "env_steps_per_second": 3995.70848615409, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.144408213658371, "eta_seconds": 18519.869569170136, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004342201401708934, "chaser_grad_norm": 0.1626635678112507, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 22.83955955505371, "collisions_per_episode": 0.5250000357627869, "entropy": 0.027943478897213936, "explorer_grad_norm": 2.787310838699341, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.70000457763672, "explorer_subspace_norm": 35.686798095703125, "final_distance": 2.20277738571167, "max_param_abs": 1.6910988092422485, "policy_loss": -0.0016410909520914174, "value_loss": 0.17956972122192383}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:53.547740+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004213.5477538, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5539804370719422, "elapsed_seconds": 1994.329573458992, "env_steps_per_second": 4011.373098241076, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.98582393364748, "eta_seconds": 17948.966161130928, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.002541680116437073, "chaser_grad_norm": 0.11740930378437042, "chaser_new_fields": 91.875, "chaser_partner_vision": 0.0, "chaser_return": 8.315001487731934, "chaser_subspace_norm": 28.9432373046875, "collisions_per_episode": 0.800000011920929, "entropy": 0.011052386136725545, "explorer_grad_norm": 2.8440057039260864, "explorer_new_fields": 94.625, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 48.87232208251953, "final_distance": 2.915606737136841, "max_param_abs": 1.8323895931243896, "policy_loss": -0.0014348829537453867, "value_loss": 0.140011016279459}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:05:06.266271+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004706.266286, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6928622715927647, "elapsed_seconds": 2494.304177733953, "env_steps_per_second": 4009.1341261693615, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8500359011493535, "eta_seconds": 17460.12924413767, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002890846548755377, "chaser_grad_norm": 0.054316350258886814, "chaser_new_fields": 92.5, "chaser_partner_vision": 0.0, "chaser_return": 8.444999694824219, "chaser_subspace_norm": 31.604827880859375, "collisions_per_episode": 0.824999988079071, "entropy": 0.008672359632328153, "explorer_grad_norm": 2.87642902135849, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 55.69298553466797, "final_distance": 2.080470323562622, "max_param_abs": 1.9203174114227295, "policy_loss": -0.0009162586192132949, "value_loss": 0.13287755846977234}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:13:26.240872+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005206.2408843, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8267417508294521, "elapsed_seconds": 2976.2703029860277, "env_steps_per_second": 4031.891857389653, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.684869921366895, "eta_seconds": 16865.53171692082, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0026825206270671975, "chaser_grad_norm": 0.3681898396462202, "chaser_new_fields": 92.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.394999504089355, "chaser_subspace_norm": 34.616058349609375, "collisions_per_episode": 0.800000011920929, "entropy": 0.011180124478414655, "explorer_grad_norm": 1.9449609816074371, "explorer_new_fields": 94.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.92500305175781, "explorer_subspace_norm": 55.717529296875, "final_distance": 2.862488269805908, "max_param_abs": 1.992835521697998, "policy_loss": -0.0040723030911706815, "value_loss": 0.17838570103049278}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:28.206999+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005688.2070158, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9603808076341456, "elapsed_seconds": 3457.370907482924, "env_steps_per_second": 4049.31966359156, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.527509521703829, "eta_seconds": 16299.034278133784, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0023234201585415892, "chaser_grad_norm": 0.109742546454072, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 35.43597412109375, "collisions_per_episode": 1.100000023841858, "entropy": 0.009827223373576999, "explorer_grad_norm": 0.8428630605340004, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.1875, "explorer_subspace_norm": 64.09198760986328, "final_distance": 2.3436453342437744, "max_param_abs": 2.070355176925659, "policy_loss": -0.0018424933510772235, "value_loss": 0.1317814216017723}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:29.307603+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006169.307615, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0946459529144341, "elapsed_seconds": 3940.7254304919625, "env_steps_per_second": 4060.166150170617, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.3785838116577365, "eta_seconds": 15762.90172196785, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00019805408916984124, "chaser_grad_norm": 0.05216287728399038, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.355000495910645, "chaser_subspace_norm": 39.39981460571289, "collisions_per_episode": 1.149999976158142, "entropy": 0.007402043207548559, "explorer_grad_norm": 1.2529384195804596, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.7125015258789, "explorer_subspace_norm": 67.51615905761719, "final_distance": 2.6458470821380615, "max_param_abs": 2.130131244659424, "policy_loss": -0.00029325868722196446, "value_loss": 0.11367498151957989}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:32.662127+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006652.6621435, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.233532814868603, "elapsed_seconds": 4440.718133526971, "env_steps_per_second": 4053.3984501519762, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.248835251214077, "eta_seconds": 15295.806904370676, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.000477152812809134, "chaser_grad_norm": 0.09721498657017946, "chaser_new_fields": 92.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.425000190734863, "chaser_subspace_norm": 41.61264419555664, "collisions_per_episode": 1.0, "entropy": 0.006108463974669576, "explorer_grad_norm": 1.4655251801013947, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 72.17704772949219, "final_distance": 2.4420201778411865, "max_param_abs": 2.179051160812378, "policy_loss": -0.0006444988547595898, "value_loss": 0.0815439410507679}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:52.654831+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007152.6548455, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.371002308455257, "elapsed_seconds": 4935.608310438925, "env_steps_per_second": 4052.185413031974, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.113006925365771, "eta_seconds": 14806.824931316776, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.008122568582032863, "chaser_grad_norm": 0.30936599150300026, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.270000457763672, "chaser_subspace_norm": 42.52458953857422, "collisions_per_episode": 1.774999976158142, "entropy": 0.008353819604963064, "explorer_grad_norm": 1.0237028822302818, "explorer_new_fields": 93.5, "explorer_partner_vision": 0.0, "explorer_return": 89.4625015258789, "explorer_subspace_norm": 77.6466293334961, "final_distance": 2.9305949211120605, "max_param_abs": 2.235593557357788, "policy_loss": -0.001684646227046116, "value_loss": 0.23588852211833}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:54:07.545006+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007647.545027, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.5076593793275082, "elapsed_seconds": 5427.573765579029, "env_steps_per_second": 4053.3765085831087, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.974738363681612, "eta_seconds": 14309.058109253803, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0010398755102261903, "chaser_grad_norm": 0.03989175520837307, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.380000114440918, "chaser_subspace_norm": 45.31040573120117, "collisions_per_episode": 1.0750000476837158, "entropy": 0.007315914612263441, "explorer_grad_norm": 0.41754402965307236, "explorer_new_fields": 94.375, "explorer_partner_vision": 0.0, "explorer_return": 91.0, "explorer_subspace_norm": 81.83265686035156, "final_distance": 2.7675256729125977, "max_param_abs": 2.2861053943634033, "policy_loss": -0.0010885612093822772, "value_loss": 0.14534751698374748}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:02:19.510460+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008139.5104733, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.647049954340554, "elapsed_seconds": 5929.379835625994, "env_steps_per_second": 4047.640843617198, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.843116560127959, "eta_seconds": 13835.219616460652, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.001582655426779822, "chaser_grad_norm": 0.08839346002787352, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.49000072479248, "chaser_subspace_norm": 49.190311431884766, "collisions_per_episode": 0.699999988079071, "entropy": 0.003298768715467304, "explorer_grad_norm": 0.4850756637752056, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 83.96345520019531, "final_distance": 2.77482008934021, "max_param_abs": 2.3289451599121094, "policy_loss": -0.001331937531009686, "value_loss": 0.06736214831471443}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:10:41.316531+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008641.3165467, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7844666381213918, "elapsed_seconds": 6424.07989723701, "env_steps_per_second": 4047.27220332091, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7061999407136597, "eta_seconds": 13342.319786569175, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.002717901549796231, "chaser_grad_norm": 0.12501308415085077, "chaser_new_fields": 92.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 48.52886199951172, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0036276100436225533, "explorer_grad_norm": 0.7959117516875267, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 85.16908264160156, "final_distance": 2.8057403564453125, "max_param_abs": 2.3636598587036133, "policy_loss": -0.0010038290538361139, "value_loss": 0.0742267556488514}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:18:56.016592+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009136.016604, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.9212862815591507, "elapsed_seconds": 6916.630613612942, "env_steps_per_second": 4048.2138723574303, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.568103094324137, "eta_seconds": 12845.171139566894, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00022324362465875325, "chaser_grad_norm": 0.024161220528185368, "chaser_new_fields": 92.875, "chaser_partner_vision": 0.0, "chaser_return": 8.52500057220459, "chaser_subspace_norm": 49.56332778930664, "collisions_per_episode": 0.5, "entropy": 0.0038169848266988993, "explorer_grad_norm": 0.2579614296555519, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 90.28097534179688, "final_distance": 2.8356008529663086, "max_param_abs": 2.406066417694092, "policy_loss": -0.000252534526921977, "value_loss": 0.05667364690452814}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:27:08.567310+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009628.567326, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.057540595391927, "elapsed_seconds": 7407.146143410937, "env_steps_per_second": 4050.1428511285208, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.429234325653212, "eta_seconds": 12345.243572351563, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.001062207278654359, "chaser_grad_norm": 0.03593328036367893, "chaser_new_fields": 92.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 51.36481475830078, "collisions_per_episode": 0.925000011920929, "entropy": 0.005192454671487212, "explorer_grad_norm": 1.0045685917139053, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 93.41552734375, "final_distance": 2.96689772605896, "max_param_abs": 2.437062978744507, "policy_loss": -0.0007657717628184457, "value_loss": 0.11875865235924721}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:35:19.082841+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010119.082855, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.195929895317208, "elapsed_seconds": 7905.347623141948, "env_steps_per_second": 4047.892834759584, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2938948429758117, "eta_seconds": 11858.021434712922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.000996393159171327, "chaser_grad_norm": 0.12287321500480175, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 51.35920333862305, "collisions_per_episode": 0.824999988079071, "entropy": 0.007126937969587743, "explorer_grad_norm": 0.5258216448128223, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 96.66453552246094, "final_distance": 2.1577203273773193, "max_param_abs": 2.472053289413452, "policy_loss": -0.0013220983569048883, "value_loss": 0.09719987586140633}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:43:37.284319+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010617.2843356, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.3308561228886053, "elapsed_seconds": 8391.082042398979, "env_steps_per_second": 4051.920816433768, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1535112250845834, "eta_seconds": 11352.6404103045, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0006369755563935264, "chaser_grad_norm": 0.15209951996803284, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 55.33511734008789, "collisions_per_episode": 1.1750000715255737, "entropy": 0.003968458739109337, "explorer_grad_norm": 0.6275946795940399, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.67500305175781, "explorer_subspace_norm": 97.49080657958984, "final_distance": 2.787330150604248, "max_param_abs": 2.507138252258301, "policy_loss": -0.000677521714706586, "value_loss": 0.0855429582297802}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:51:43.018737+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011103.0187495, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.467534301508341, "elapsed_seconds": 8883.123485430027, "env_steps_per_second": 4052.6285668601463, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.0158752573990837, "eta_seconds": 10857.150926636701, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 6.216459688701548e-05, "chaser_grad_norm": 0.0843466967344284, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.745000839233398, "chaser_subspace_norm": 53.2131233215332, "collisions_per_episode": 0.7250000238418579, "entropy": 0.007430837024003267, "explorer_grad_norm": 0.35458748042583466, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 98.6747817993164, "final_distance": 2.9391210079193115, "max_param_abs": 2.5464768409729004, "policy_loss": -0.0005818257509364599, "value_loss": 0.08211887069046497}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:59:55.060182+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011595.0601966, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.604981942926954, "elapsed_seconds": 9377.934994537034, "env_steps_per_second": 4052.0647692841007, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.879190568498212, "eta_seconds": 10365.086046593564, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0010052093184430566, "chaser_grad_norm": 0.09301582723855972, "chaser_new_fields": 94.125, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 56.764808654785156, "collisions_per_episode": 0.949999988079071, "entropy": 0.0033797190408222377, "explorer_grad_norm": 0.5405961982905865, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.5625, "explorer_subspace_norm": 102.1544189453125, "final_distance": 2.845025062561035, "max_param_abs": 2.579113483428955, "policy_loss": -0.0005198264389183471, "value_loss": 0.0823406670242548}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:08:09.871691+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012089.8717058, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.7412828842372012, "elapsed_seconds": 9868.618383253925, "env_steps_per_second": 4053.2522838127034, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7412828842372012, "eta_seconds": 9868.618383253925, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0023042803507682397, "chaser_grad_norm": 0.04860458383336663, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 59.11008834838867, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0030235633021220565, "explorer_grad_norm": 0.4425838142633438, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 102.14756774902344, "final_distance": 2.8141345977783203, "max_param_abs": 2.6225476264953613, "policy_loss": -0.0009095669351530944, "value_loss": 0.09786444157361984}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:16:20.555079+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012580.555093, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.879656619975544, "elapsed_seconds": 10366.763831911958, "env_steps_per_second": 4051.4089720758957, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.605403608549302, "eta_seconds": 9379.452990777487, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005118195990925223, "chaser_grad_norm": 0.08334831520915031, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 59.05033874511719, "collisions_per_episode": 0.800000011920929, "entropy": 0.003657435707282275, "explorer_grad_norm": 0.2816202864050865, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 106.18612670898438, "final_distance": 2.3185527324676514, "max_param_abs": 2.6511926651000977, "policy_loss": -0.0004923692047587025, "value_loss": 0.07210455648601055}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:24:38.700526+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013078.7005394, "update": 10500}
+{"device": "cuda", "elapsed_hours": 3.0160725797041477, "elapsed_seconds": 10857.86128693493, "env_steps_per_second": 4052.363429338005, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.467695747030666, "eta_seconds": 8883.704689310398, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00017792440830510792, "chaser_grad_norm": 0.04437321983277798, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 57.75242233276367, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0032481419038958848, "explorer_grad_norm": 0.24263937771320343, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 102.43596649169922, "final_distance": 2.4457814693450928, "max_param_abs": 2.6756579875946045, "policy_loss": -0.00031847636937154533, "value_loss": 0.06373031809926033}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:32:49.797984+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013569.797999, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.155583950798868, "elapsed_seconds": 11360.102222875925, "env_steps_per_second": 4049.259337417708, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3323881375469893, "eta_seconds": 8396.597295169162, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.008585023168623529, "chaser_grad_norm": 0.234678759239614, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 58.88050079345703, "collisions_per_episode": 0.6000000238418579, "entropy": 0.005969225661829114, "explorer_grad_norm": 0.48086490482091904, "explorer_new_fields": 94.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.63750457763672, "explorer_subspace_norm": 100.81809997558594, "final_distance": 2.4236786365509033, "max_param_abs": 2.7097814083099365, "policy_loss": -0.004882263489065508, "value_loss": 0.20253624022006989}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:41:12.038918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014072.0389283, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.289983558591662, "elapsed_seconds": 11843.940810929984, "env_steps_per_second": 4052.7051566910905, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1933223723944413, "eta_seconds": 7895.960540619989, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00019440097922895916, "chaser_grad_norm": 0.029068120289593935, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.894999504089355, "chaser_subspace_norm": 62.05281448364258, "collisions_per_episode": 0.625, "entropy": 0.003293231420684606, "explorer_grad_norm": 0.493771456182003, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 110.28379821777344, "final_distance": 2.185750961303711, "max_param_abs": 2.739060878753662, "policy_loss": -0.0002078625110115695, "value_loss": 0.06744680739939213}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:49:15.877507+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014555.8775215, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.4264354140583437, "elapsed_seconds": 12335.167490610038, "env_steps_per_second": 4053.4512432086353, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0558612484350065, "eta_seconds": 7401.100494366023, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005562198738733315, "chaser_grad_norm": 0.08132471237331629, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.86500072479248, "chaser_subspace_norm": 60.91058349609375, "collisions_per_episode": 0.625, "entropy": 0.0040358962723985314, "explorer_grad_norm": 0.3306007981300354, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 111.05867767333984, "final_distance": 2.458824634552002, "max_param_abs": 2.7716782093048096, "policy_loss": -0.0005716206628125775, "value_loss": 0.06918930821120739}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:57:27.104189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015047.104203, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.563489204867219, "elapsed_seconds": 12828.561137521989, "env_steps_per_second": 4053.4553674851572, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9188018795438873, "eta_seconds": 6907.686766357994, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00011050143215055735, "chaser_grad_norm": 0.049098674207925797, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.920000076293945, "chaser_subspace_norm": 63.42983627319336, "collisions_per_episode": 0.699999988079071, "entropy": 0.003923855605535209, "explorer_grad_norm": 0.3020031377673149, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 110.25255584716797, "final_distance": 2.292327404022217, "max_param_abs": 2.8057360649108887, "policy_loss": -0.0004508834210561474, "value_loss": 0.04780322127044201}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:05:40.497834+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015540.4978487, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.69906137363028, "elapsed_seconds": 13316.620945069008, "env_steps_per_second": 4055.0827588131942, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.781029550266431, "eta_seconds": 6411.706380959152, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00023009996807599187, "chaser_grad_norm": 0.06983139086514711, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.8100004196167, "chaser_subspace_norm": 63.69921112060547, "collisions_per_episode": 0.699999988079071, "entropy": 0.003970191464759409, "explorer_grad_norm": 0.6755344867706299, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 115.0482406616211, "final_distance": 2.1381518840789795, "max_param_abs": 2.828923225402832, "policy_loss": -0.0004508252339432417, "value_loss": 0.06965262070298195}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:13:48.557642+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016028.557656, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.835262262045257, "elapsed_seconds": 13806.944143362925, "env_steps_per_second": 4055.930075368597, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6436838265908245, "eta_seconds": 5917.261775726968, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 7.470461521874583e-05, "chaser_grad_norm": 0.07316626608371735, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 67.3201675415039, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0024352622567676008, "explorer_grad_norm": 0.40397659689188004, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 119.69538116455078, "final_distance": 2.3341729640960693, "max_param_abs": 2.8383781909942627, "policy_loss": -7.460785411250725e-05, "value_loss": 0.09181175008416176}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:21:58.880839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016518.880855, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.970110981043318, "elapsed_seconds": 14292.399531755946, "env_steps_per_second": 4058.100941771965, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5059041652233276, "eta_seconds": 5421.25499480398, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 2.7242026847673628e-05, "chaser_grad_norm": 0.0497783487662673, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 67.51028442382812, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0024516364210285246, "explorer_grad_norm": 0.25298214703798294, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 121.70927429199219, "final_distance": 2.5439205169677734, "max_param_abs": 2.869907855987549, "policy_loss": -6.890869436126934e-05, "value_loss": 0.05644155852496624}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:30:04.336228+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017004.3362422, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.107270685498596, "elapsed_seconds": 14786.174467794946, "env_steps_per_second": 4057.844720463911, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3690902284995319, "eta_seconds": 4928.724822598315, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0006331868479183891, "chaser_grad_norm": 0.04258312797173858, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.51932525634766, "collisions_per_episode": 0.375, "entropy": 0.0030604162020608783, "explorer_grad_norm": 0.4632948637008667, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 122.46908569335938, "final_distance": 2.23240065574646, "max_param_abs": 2.899177074432373, "policy_loss": -0.00021183396170698643, "value_loss": 0.04441228974610567}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:38:18.111163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017498.1111782, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.24289645000694, "elapsed_seconds": 15274.427220024983, "env_steps_per_second": 4059.072010158073, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2318086467762082, "eta_seconds": 4434.51112839435, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0013615168079588735, "chaser_grad_norm": 0.04086262080818415, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.875, "chaser_subspace_norm": 69.30021667480469, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0028082082280889153, "explorer_grad_norm": 0.1993851512670517, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.7249984741211, "explorer_subspace_norm": 119.8309555053711, "final_distance": 2.0393521785736084, "max_param_abs": 2.918504238128662, "policy_loss": -0.00019859792260112918, "value_loss": 0.04485477786511183}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:46:26.363915+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017986.3639271, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.377337626529431, "elapsed_seconds": 15758.41545550595, "env_steps_per_second": 4061.322039687598, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0943344066323577, "eta_seconds": 3939.6038638764876, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006552253194327395, "chaser_grad_norm": 0.026540640741586685, "chaser_new_fields": 94.75, "chaser_partner_vision": 0.0, "chaser_return": 8.905000686645508, "chaser_subspace_norm": 69.47985076904297, "collisions_per_episode": 0.45000001788139343, "entropy": 0.004001817433163524, "explorer_grad_norm": 0.2520532086491585, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.5374984741211, "explorer_subspace_norm": 132.6637725830078, "final_distance": 2.078486680984497, "max_param_abs": 2.937121629714966, "policy_loss": -0.0010853266977051135, "value_loss": 0.08242643252015114}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:54:30.352153+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018470.3521686, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.5111880817922065, "elapsed_seconds": 16240.277094451943, "env_steps_per_second": 4063.970067514866, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9569186840165287, "eta_seconds": 3444.9072624595033, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0011654459594116588, "chaser_grad_norm": 0.12811908777803183, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 72.75859069824219, "collisions_per_episode": 1.475000023841858, "entropy": 0.0036797402426600456, "explorer_grad_norm": 0.2041437365114689, "explorer_new_fields": 94.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 129.2154998779297, "final_distance": 2.415665864944458, "max_param_abs": 2.957319498062134, "policy_loss": -0.0006939259038780676, "value_loss": 0.2137848101556301}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:02:32.213790+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018952.2138023, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.6476044810977575, "elapsed_seconds": 16731.376131951925, "env_steps_per_second": 4064.220388312252, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.82016549666431, "eta_seconds": 2952.5957879915163, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 5.5648104003269694e-05, "chaser_grad_norm": 0.018644486088305712, "chaser_new_fields": 94.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.880000114440918, "chaser_subspace_norm": 72.43238067626953, "collisions_per_episode": 0.7250000238418579, "entropy": 0.001984165806788951, "explorer_grad_norm": 0.2640278488397598, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 126.32189178466797, "final_distance": 2.4246740341186523, "max_param_abs": 2.9786489009857178, "policy_loss": -5.09057059243645e-05, "value_loss": 0.043034049682319164}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:10:43.312827+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019443.3128421, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.785597953629169, "elapsed_seconds": 17228.15263306501, "env_steps_per_second": 4063.117009170966, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6836568505184527, "eta_seconds": 2461.16466186643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 2.6341494720749647e-05, "chaser_grad_norm": 0.07059159502387047, "chaser_new_fields": 94.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.920001029968262, "chaser_subspace_norm": 75.26416778564453, "collisions_per_episode": 0.42500001192092896, "entropy": 0.002113209222443402, "explorer_grad_norm": 0.316756010055542, "explorer_new_fields": 95.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 127.33657836914062, "final_distance": 2.3594603538513184, "max_param_abs": 2.989992141723633, "policy_loss": -4.8084022527206116e-05, "value_loss": 0.026043227408081293}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:19:00.089328+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019940.089339, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.921170455568612, "elapsed_seconds": 17716.213640047004, "env_steps_per_second": 4064.0738175140323, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5467967172854014, "eta_seconds": 1968.468182227445, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0008898909904117291, "chaser_grad_norm": 0.042340966407209635, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 72.6676254272461, "collisions_per_episode": 0.800000011920929, "entropy": 0.001199792983243242, "explorer_grad_norm": 0.21852006390690804, "explorer_new_fields": 95.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 124.67669677734375, "final_distance": 2.4179787635803223, "max_param_abs": 3.023815155029297, "policy_loss": -0.0005869875105570799, "value_loss": 0.08416097238659859}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:27:08.150335+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020428.1503508, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.057042567851927, "elapsed_seconds": 18205.35324426694, "env_steps_per_second": 4064.7384869229822, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4100304784744806, "eta_seconds": 1476.10972250813, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 9.724407010380798e-05, "chaser_grad_norm": 0.053080628626048565, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.905000686645508, "chaser_subspace_norm": 78.34722137451172, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0021247865515761077, "explorer_grad_norm": 0.20850447192788124, "explorer_new_fields": 95.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 131.81642150878906, "final_distance": 2.267756223678589, "max_param_abs": 3.0450665950775146, "policy_loss": -0.0002477926192216984, "value_loss": 0.03876691497862339}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:35:17.289939+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020917.289952, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.1930706142980485, "elapsed_seconds": 18695.054211472976, "env_steps_per_second": 4065.2463020599066, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2733195060156868, "eta_seconds": 983.9502216564724, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0028486762956183576, "chaser_grad_norm": 0.13814295083284378, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 75.9752197265625, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0027867945027537644, "explorer_grad_norm": 0.16607102751731873, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 136.76522827148438, "final_distance": 2.363304853439331, "max_param_abs": 3.070605754852295, "policy_loss": -0.0013202877221658582, "value_loss": 0.09105210937559605}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:43:26.990907+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021406.990922, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.327256802463316, "elapsed_seconds": 19178.124488867936, "env_steps_per_second": 4067.1338871157914, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13659632826829013, "eta_seconds": 491.7467817658445, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0014851364632712527, "chaser_grad_norm": 0.08397708088159561, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 80.55755615234375, "collisions_per_episode": 0.800000011920929, "entropy": 0.0034964141086675227, "explorer_grad_norm": 0.2840192373842001, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 135.56906127929688, "final_distance": 2.4301421642303467, "max_param_abs": 3.0977866649627686, "policy_loss": -0.0005637188296532258, "value_loss": 0.09781094640493393}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:51:30.061184+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021890.0611954, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.456585522624987, "elapsed_seconds": 19643.707881449955, "env_steps_per_second": 4072.550889210993, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:59:15.644576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022355.6445875, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.457300321458897, "elapsed_seconds": 19646.28115725203, "env_steps_per_second": 4072.0174652733, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:59:18.217854+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022358.2178695, "update": 20000}

non_social/seed_0001/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.457300321458897,
+  "elapsed_seconds": 19646.28115725203,
+  "env_steps_per_second": 4072.0174652733,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 0.0031268233263772593,
+    "chaser_grad_norm": 0.2260504849255085,
+    "chaser_new_fields": 93.4000015258789,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.640000343322754,
+    "chaser_subspace_norm": 81.0290298461914,
+    "collisions_per_episode": 1.149999976158142,
+    "entropy": 0.003422852954827249,
+    "explorer_grad_norm": 0.24985111132264137,
+    "explorer_new_fields": 94.82500457763672,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 91.82500457763672,
+    "explorer_subspace_norm": 135.97732543945312,
+    "final_distance": 2.2879505157470703,
+    "max_param_abs": 3.1244027614593506,
+    "policy_loss": -0.0035095140663869984,
+    "value_loss": 0.2153652273118496
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-02T19:59:18.217854+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783022358.2178695,
+  "update": 20000
+}

non_social/seed_0002/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b05e3f4ca6b8f514ca68a393b1bdc046876c2fdb9517fb51b9942361d3a57ac
+size 29760368

non_social/seed_0002/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01",
+  "seed": 2,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0002/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0007321728369182286, "elapsed_seconds": 2.635822212905623, "env_steps_per_second": 1517.5530354115056, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.642724565527654, "eta_seconds": 52713.808435899555, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0061279205385336155, "chaser_grad_norm": 8.69063800573349, "chaser_new_fields": 32.35000228881836, "chaser_partner_vision": 0.0, "chaser_return": -3.5600006580352783, "chaser_subspace_norm": 0.034794267266988754, "collisions_per_episode": 1.6750000715255737, "entropy": 2.7557111382484436, "explorer_grad_norm": 6.052844345569611, "explorer_new_fields": 35.82500076293945, "explorer_partner_vision": 0.0, "explorer_return": 3.4375, "explorer_subspace_norm": 0.031101124361157417, "final_distance": 5.287587642669678, "max_param_abs": 1.0012022256851196, "policy_loss": -0.01202418064700339, "value_loss": 7.6772836446762085}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.814566+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.8145802, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.13635041027666173, "elapsed_seconds": 490.86147699598223, "env_steps_per_second": 4074.4692621628774, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.317666000789807, "eta_seconds": 19143.597602843307, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0014374513662449573, "chaser_grad_norm": 0.22422203421592712, "chaser_new_fields": 90.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.045001029968262, "chaser_subspace_norm": 11.34537124633789, "collisions_per_episode": 0.949999988079071, "entropy": 0.07179526425898075, "explorer_grad_norm": 10.016987323760986, "explorer_new_fields": 92.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 88.1875, "explorer_subspace_norm": 16.648193359375, "final_distance": 2.3869614601135254, "max_param_abs": 1.259480357170105, "policy_loss": -0.0028794059618419965, "value_loss": 0.48020658642053604}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:03.040222+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003203.0402334, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.27326310440720086, "elapsed_seconds": 983.747175865923, "env_steps_per_second": 4066.085370439903, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.191998983736816, "eta_seconds": 18691.196341452538, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0014687455615272071, "chaser_grad_norm": 0.20035238936543465, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 14.68276309967041, "collisions_per_episode": 1.75, "entropy": 0.028629083652049303, "explorer_grad_norm": 5.136371970176697, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 90.8125, "explorer_subspace_norm": 29.166044235229492, "final_distance": 2.825841188430786, "max_param_abs": 1.3679320812225342, "policy_loss": -0.0019738255795118675, "value_loss": 0.426460824906826}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:15.925921+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003695.9259365, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.40687836955388773, "elapsed_seconds": 1464.762130393996, "env_steps_per_second": 4096.228237677133, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.018166557831282, "eta_seconds": 18065.399608192616, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0012405988440074012, "chaser_grad_norm": 0.11184354312717915, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.59000015258789, "chaser_subspace_norm": 16.197612762451172, "collisions_per_episode": 1.5750000476837158, "entropy": 0.020140967797487974, "explorer_grad_norm": 4.171184957027435, "explorer_new_fields": 96.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.13750457763672, "explorer_subspace_norm": 39.99115753173828, "final_distance": 2.9582443237304688, "max_param_abs": 1.439399003982544, "policy_loss": -0.0013640758586177526, "value_loss": 0.2158144824206829}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:16.940873+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004176.9408846, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5437295976183183, "elapsed_seconds": 1957.426551425946, "env_steps_per_second": 4086.9988169784256, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.893566378564865, "eta_seconds": 17616.838962833514, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0013608692952107315, "chaser_grad_norm": 0.10284747928380966, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 18.577129364013672, "collisions_per_episode": 1.850000023841858, "entropy": 0.009836691664531827, "explorer_grad_norm": 2.7826343774795532, "explorer_new_fields": 96.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.4375, "explorer_subspace_norm": 51.28536605834961, "final_distance": 2.7944366931915283, "max_param_abs": 1.4927884340286255, "policy_loss": -0.0007597444546405541, "value_loss": 0.19846593216061592}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:29.605296+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004669.6053092, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6781449054674724, "elapsed_seconds": 2441.3216596829006, "env_steps_per_second": 4096.141923919556, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.747014338272307, "eta_seconds": 17089.251617780305, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00016304003913525378, "chaser_grad_norm": 0.0689720269292593, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.545000076293945, "chaser_subspace_norm": 21.22562026977539, "collisions_per_episode": 2.049999952316284, "entropy": 0.011092229979112744, "explorer_grad_norm": 3.2772421091794968, "explorer_new_fields": 96.625, "explorer_partner_vision": 0.0, "explorer_return": 93.70000457763672, "explorer_subspace_norm": 56.04767608642578, "final_distance": 2.6179006099700928, "max_param_abs": 1.5319234132766724, "policy_loss": -0.00040609817267522885, "value_loss": 0.287984736263752}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:33.500402+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005153.5004132, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8128366362458069, "elapsed_seconds": 2926.211890484905, "env_steps_per_second": 4100.865025878721, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.606074272059572, "eta_seconds": 16581.86737941446, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0006030650421495198, "chaser_grad_norm": 0.07318039052188396, "chaser_new_fields": 93.75, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 24.370473861694336, "collisions_per_episode": 1.5500000715255737, "entropy": 0.006501625059172511, "explorer_grad_norm": 1.3841366171836853, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.07500457763672, "explorer_subspace_norm": 62.4726676940918, "final_distance": 3.1512153148651123, "max_param_abs": 1.56986665725708, "policy_loss": -0.0006041124149902544, "value_loss": 0.20404154062271118}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:38.390636+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005638.3906512, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9494057591094235, "elapsed_seconds": 3417.8607327939244, "env_steps_per_second": 4096.129448947946, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.475770007230139, "eta_seconds": 16112.7720260285, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 5.0752096209590036e-05, "chaser_grad_norm": 0.023512616753578186, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.580000877380371, "chaser_subspace_norm": 25.51074981689453, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007038616458885372, "explorer_grad_norm": 1.4274124652147293, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 66.74330139160156, "final_distance": 2.7487001419067383, "max_param_abs": 1.606393575668335, "policy_loss": -0.0008341384197114721, "value_loss": 0.16259713098406792}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:50.039478+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006130.039492, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0876443207569213, "elapsed_seconds": 3915.519554724917, "env_steps_per_second": 4086.3031780016418, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.350577283027685, "eta_seconds": 15662.078218899667, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0017131934787162212, "chaser_grad_norm": 0.0729985972866416, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 26.832889556884766, "collisions_per_episode": 1.75, "entropy": 0.008575530722737312, "explorer_grad_norm": 1.4874772131443024, "explorer_new_fields": 96.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 71.17113494873047, "final_distance": 3.008681535720825, "max_param_abs": 1.636663556098938, "policy_loss": -0.0019146683402411213, "value_loss": 0.19464126974344254}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:07.698298+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006627.698313, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.2238280326572002, "elapsed_seconds": 4405.78091756592, "env_steps_per_second": 4085.541323272273, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.2154076680414665, "eta_seconds": 15175.467604949281, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0012005009262245625, "chaser_grad_norm": 0.11612079106271267, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.645001411437988, "chaser_subspace_norm": 27.53525733947754, "collisions_per_episode": 1.600000023841858, "entropy": 0.004619026556611061, "explorer_grad_norm": 1.6260186731815338, "explorer_new_fields": 96.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.07500457763672, "explorer_subspace_norm": 77.10869598388672, "final_distance": 2.85575795173645, "max_param_abs": 1.6697918176651, "policy_loss": -0.0011311787646235416, "value_loss": 0.18725960701704025}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:17.959662+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007117.9596765, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.3576553877099973, "elapsed_seconds": 4887.55939575599, "env_steps_per_second": 4092.0218826121236, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.072966163129992, "eta_seconds": 14662.678187267971, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005470263573985612, "chaser_grad_norm": 0.08314792811870575, "chaser_new_fields": 93.0, "chaser_partner_vision": 0.0, "chaser_return": 8.460000038146973, "chaser_subspace_norm": 30.84358787536621, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007231791387312114, "explorer_grad_norm": 2.956935614347458, "explorer_new_fields": 96.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 82.12212371826172, "final_distance": 2.6573290824890137, "max_param_abs": 1.700556755065918, "policy_loss": -0.0010330596112169133, "value_loss": 0.18062416464090347}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:19.738140+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007599.7381537, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.4912110790266564, "elapsed_seconds": 5368.359884495963, "env_steps_per_second": 4098.085909541362, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.931374662888458, "eta_seconds": 14152.948786398449, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.001206373677458461, "chaser_grad_norm": 0.08986116107553244, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 32.25676727294922, "collisions_per_episode": 1.4250000715255737, "entropy": 0.0041676408145576715, "explorer_grad_norm": 0.5372973829507828, "explorer_new_fields": 96.75, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 85.73692321777344, "final_distance": 2.9968020915985107, "max_param_abs": 1.729843020439148, "policy_loss": -0.0007324595502566922, "value_loss": 0.14542652666568756}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:20.538629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008080.5386395, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.625052962143317, "elapsed_seconds": 5850.190663715941, "env_steps_per_second": 4102.430395790829, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.791790245001073, "eta_seconds": 13650.444882003863, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006094975457799912, "chaser_grad_norm": 0.05917355325073004, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 33.52567672729492, "collisions_per_episode": 1.5, "entropy": 0.00545320357196033, "explorer_grad_norm": 0.6449002400040627, "explorer_new_fields": 96.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.5999984741211, "explorer_subspace_norm": 85.36280822753906, "final_distance": 2.549459934234619, "max_param_abs": 1.7547281980514526, "policy_loss": -0.0006153682163017038, "value_loss": 0.12353493086993694}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:22.369408+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008562.3694234, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7589124985274975, "elapsed_seconds": 6332.084994698991, "env_steps_per_second": 4106.072489830179, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.653125958480187, "eta_seconds": 13151.253450528673, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00042492411724981416, "chaser_grad_norm": 0.04196750558912754, "chaser_new_fields": 94.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.745000839233398, "chaser_subspace_norm": 35.55250930786133, "collisions_per_episode": 1.0500000715255737, "entropy": 0.004787102923728526, "explorer_grad_norm": 1.0047070682048798, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 89.75172424316406, "final_distance": 3.230062246322632, "max_param_abs": 1.7901263236999512, "policy_loss": -0.000798380141378896, "value_loss": 0.10945381037890911}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:24.263738+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009044.263749, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.8928185786483098, "elapsed_seconds": 6814.146883133915, "env_steps_per_second": 4109.098391950488, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5152345032040038, "eta_seconds": 12654.844211534413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00019714361857248175, "chaser_grad_norm": 0.03700845502316952, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 36.750267028808594, "collisions_per_episode": 1.600000023841858, "entropy": 0.0030598515877500176, "explorer_grad_norm": 1.1551428139209747, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.63750457763672, "explorer_subspace_norm": 88.27851104736328, "final_distance": 3.290786027908325, "max_param_abs": 1.8184542655944824, "policy_loss": -0.0002503604961097139, "value_loss": 0.1453133225440979}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:26.325628+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009526.3256426, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.0287750570827563, "elapsed_seconds": 7303.590205497923, "env_steps_per_second": 4107.568901855542, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.381291761804594, "eta_seconds": 12172.65034249654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0006325586149842077, "chaser_grad_norm": 0.07497241348028183, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 38.45086669921875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.004371156450361013, "explorer_grad_norm": 1.1777122020721436, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 98.05054473876953, "final_distance": 2.680711030960083, "max_param_abs": 1.8433010578155518, "policy_loss": -0.0005633578459964816, "value_loss": 0.09415272064507008}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:35.768948+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010015.7689602, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.165660841056637, "elapsed_seconds": 7796.379027803894, "env_steps_per_second": 4104.469508970737, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.248491261584956, "eta_seconds": 11694.568541705841, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00016110613666535205, "chaser_grad_norm": 0.05154063366353512, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 38.86147689819336, "collisions_per_episode": 0.925000011920929, "entropy": 0.004242159542627633, "explorer_grad_norm": 1.4051325917243958, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 96.03810119628906, "final_distance": 3.4580161571502686, "max_param_abs": 1.8778479099273682, "policy_loss": -0.000521032350832229, "value_loss": 0.09992639347910881}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:41:48.557774+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010508.5577881, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.3019769951799796, "elapsed_seconds": 8287.117182647926, "env_steps_per_second": 4102.753617529542, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1144394640670305, "eta_seconds": 11211.98207064131, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00018751345209144787, "chaser_grad_norm": 0.04086209274828434, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.74000072479248, "chaser_subspace_norm": 41.277713775634766, "collisions_per_episode": 1.350000023841858, "entropy": 0.00669125746935606, "explorer_grad_norm": 0.7252416163682938, "explorer_new_fields": 96.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.5250015258789, "explorer_subspace_norm": 99.34879302978516, "final_distance": 2.824924945831299, "max_param_abs": 1.9064655303955078, "policy_loss": -0.0007197263473628368, "value_loss": 0.11501258239150047}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:49:59.295926+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010999.295938, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.439056712242212, "elapsed_seconds": 8780.604164071963, "env_steps_per_second": 4099.9456674408575, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9810693149627037, "eta_seconds": 10731.849533865734, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0008318895781759261, "chaser_grad_norm": 0.029094298603013158, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 41.468994140625, "collisions_per_episode": 1.274999976158142, "entropy": 0.003884760953951627, "explorer_grad_norm": 0.7414844408631325, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.50614929199219, "final_distance": 3.1373450756073, "max_param_abs": 1.942590355873108, "policy_loss": -0.0003571343062773735, "value_loss": 0.12418003007769585}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:12.782910+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011492.7829242, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.573374887945829, "elapsed_seconds": 9264.149596604984, "env_steps_per_second": 4101.833590200852, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8442564550980216, "eta_seconds": 10239.323238352878, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 2.222422159698506e-05, "chaser_grad_norm": 0.02208172483369708, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 44.127201080322266, "collisions_per_episode": 1.25, "entropy": 0.0022298935218714178, "explorer_grad_norm": 0.3783605508506298, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 100.45281219482422, "final_distance": 3.419586181640625, "max_param_abs": 1.9667505025863647, "policy_loss": -4.9121858756162595e-05, "value_loss": 0.10526370443403721}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:16.328340+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011976.3283513, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.7105219278494284, "elapsed_seconds": 9757.878940257942, "env_steps_per_second": 4099.25151202405, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7105219278494284, "eta_seconds": 9757.878940257942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 9.216201643482552e-06, "chaser_grad_norm": 0.03743532067164779, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 45.39440155029297, "collisions_per_episode": 1.225000023841858, "entropy": 0.002705720718950033, "explorer_grad_norm": 0.6737443953752518, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.60910034179688, "final_distance": 3.6136252880096436, "max_param_abs": 1.9912539720535278, "policy_loss": -0.00025387931091458427, "value_loss": 0.11213633045554161}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:30.057687+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012470.0577025, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.8467568718997063, "elapsed_seconds": 10248.324738838943, "env_steps_per_second": 4098.230790914446, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.57563716981402, "eta_seconds": 9272.293811330472, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00021304979432612559, "chaser_grad_norm": 0.04589147632941604, "chaser_new_fields": 94.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 42.61776351928711, "collisions_per_episode": 1.5, "entropy": 0.004361974424682558, "explorer_grad_norm": 0.6482382416725159, "explorer_new_fields": 96.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.48750305175781, "explorer_subspace_norm": 104.19003295898438, "final_distance": 3.3916919231414795, "max_param_abs": 2.0091161727905273, "policy_loss": -0.0011383400935436327, "value_loss": 0.17632096633315086}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:22:40.503482+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012960.5034928, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.9841719255210934, "elapsed_seconds": 10743.018931875937, "env_steps_per_second": 4095.6829992588273, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4415952117899855, "eta_seconds": 8789.742762443948, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 1.509851393197148e-05, "chaser_grad_norm": 0.03615148039534688, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.780000686645508, "chaser_subspace_norm": 43.497840881347656, "collisions_per_episode": 1.149999976158142, "entropy": 0.0033954427344724536, "explorer_grad_norm": 0.25947869941592216, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.66839599609375, "final_distance": 3.3145911693573, "max_param_abs": 2.024752378463745, "policy_loss": -3.67707628541325e-05, "value_loss": 0.13062887638807297}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:30:55.197678+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013455.1976929, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.1180544764244242, "elapsed_seconds": 11224.996115127928, "env_steps_per_second": 4097.996963937101, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3046489608354443, "eta_seconds": 8296.736259007599, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0022483308288680365, "chaser_grad_norm": 0.039612220134586096, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 45.17502975463867, "collisions_per_episode": 1.125, "entropy": 0.0031940711778588593, "explorer_grad_norm": 0.32754963636398315, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 108.3421859741211, "final_distance": 3.448914051055908, "max_param_abs": 2.0493438243865967, "policy_loss": -0.001800878806509587, "value_loss": 0.1461172215640545}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:38:57.174859+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013937.1748703, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.2521516086605455, "elapsed_seconds": 11707.745791177964, "env_steps_per_second": 4099.849864879115, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1681010724403635, "eta_seconds": 7805.163860785309, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.000790681346329869, "chaser_grad_norm": 0.04079975187778473, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 46.499290466308594, "collisions_per_episode": 1.0750000476837158, "entropy": 0.004843202885240316, "explorer_grad_norm": 0.3729347884654999, "explorer_new_fields": 97.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 111.02171325683594, "final_distance": 3.2637646198272705, "max_param_abs": 2.065484046936035, "policy_loss": -0.0003173275126648889, "value_loss": 0.08790707774460316}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:46:59.924537+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014419.9245532, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.3880419566197086, "elapsed_seconds": 12196.951043830952, "env_steps_per_second": 4099.385151282484, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0328251739718253, "eta_seconds": 7318.170626298572, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.003345426901507409, "chaser_grad_norm": 0.08309758501127362, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 46.250267028808594, "collisions_per_episode": 1.0500000715255737, "entropy": 0.003117095388006419, "explorer_grad_norm": 0.24165698513388634, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 111.1562271118164, "final_distance": 3.4718072414398193, "max_param_abs": 2.08314847946167, "policy_loss": -0.00027372838303563185, "value_loss": 0.08545360900461674}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:09.129788+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014909.1297987, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5235442068249943, "elapsed_seconds": 12684.75914456998, "env_steps_per_second": 4099.407754404218, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8972930344442278, "eta_seconds": 6830.25492399922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 6.795539224679813e-05, "chaser_grad_norm": 0.022474604193121195, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 46.87422561645508, "collisions_per_episode": 0.949999988079071, "entropy": 0.0026716027641668916, "explorer_grad_norm": 0.40012921392917633, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 113.71060180664062, "final_distance": 3.3464126586914062, "max_param_abs": 2.1162638664245605, "policy_loss": -0.00010689926428320007, "value_loss": 0.0879583116620779}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:16.937891+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015396.9379056, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.659062793856946, "elapsed_seconds": 13172.626057885005, "env_steps_per_second": 4099.410380489479, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.761770974820011, "eta_seconds": 6342.375509352039, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0037481113682407818, "chaser_grad_norm": 0.04439642280340195, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.755000114440918, "chaser_subspace_norm": 46.802284240722656, "collisions_per_episode": 1.4500000476837158, "entropy": 0.0057298337342217565, "explorer_grad_norm": 0.5554292425513268, "explorer_new_fields": 96.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.73750305175781, "explorer_subspace_norm": 119.93209075927734, "final_distance": 3.259939670562744, "max_param_abs": 2.131298303604126, "policy_loss": -0.0031371804652593482, "value_loss": 0.17059676349163055}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:24.804803+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015884.8048182, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.7951653686480453, "elapsed_seconds": 13662.595327132964, "env_steps_per_second": 4098.782014628501, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6264994437063054, "eta_seconds": 5855.3979973426995, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0010495001250528935, "chaser_grad_norm": 0.025360917206853628, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 49.11723709106445, "collisions_per_episode": 1.2000000476837158, "entropy": 0.0034450897946953773, "explorer_grad_norm": 0.38084179908037186, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.9749984741211, "explorer_subspace_norm": 121.4040756225586, "final_distance": 3.168457508087158, "max_param_abs": 2.159163475036621, "policy_loss": -0.001656242737077207, "value_loss": 0.12334729917347431}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:19:34.774071+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016374.774086, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.932483092390531, "elapsed_seconds": 14156.939132605912, "env_steps_per_second": 4096.930802394695, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.491631517803305, "eta_seconds": 5369.873464091898, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005145524231314202, "chaser_grad_norm": 0.021116286981850863, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.860000610351562, "chaser_subspace_norm": 52.13620376586914, "collisions_per_episode": 1.125, "entropy": 0.002387709857430309, "explorer_grad_norm": 0.2525097243487835, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.125, "explorer_subspace_norm": 119.46451568603516, "final_distance": 3.4527671337127686, "max_param_abs": 2.180997610092163, "policy_loss": -0.00031888963405890536, "value_loss": 0.07078775204718113}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:27:49.117876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016869.117887, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.066998258098611, "elapsed_seconds": 14641.193729154998, "env_steps_per_second": 4098.026507259585, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3556660860328704, "eta_seconds": 4880.397909718334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005770839916344472, "chaser_grad_norm": 0.07790196500718594, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 51.55928421020508, "collisions_per_episode": 0.7250000238418579, "entropy": 0.004281870438717306, "explorer_grad_norm": 1.035013571381569, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 116.98993682861328, "final_distance": 3.6269309520721436, "max_param_abs": 2.206885576248169, "policy_loss": -0.0004720029999729647, "value_loss": 0.06566426623612642}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:35:53.372473+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017353.3724883, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.201983206083306, "elapsed_seconds": 15127.139541899902, "env_steps_per_second": 4098.593777645095, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.219930608217734, "eta_seconds": 4391.750189583842, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0011227785522094358, "chaser_grad_norm": 0.06945330277085304, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 51.186004638671875, "collisions_per_episode": 1.2000000476837158, "entropy": 0.00486322829965502, "explorer_grad_norm": 0.4907003194093704, "explorer_new_fields": 97.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.01250457763672, "explorer_subspace_norm": 120.90361022949219, "final_distance": 3.6959824562072754, "max_param_abs": 2.233208656311035, "policy_loss": -0.000521478934562225, "value_loss": 0.09477586857974529}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:43:59.318285+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017839.3182957, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.341138527599138, "elapsed_seconds": 15628.098699356895, "env_steps_per_second": 4095.187855617628, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0852846318997844, "eta_seconds": 3907.0246748392237, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 4.187869177676301e-05, "chaser_grad_norm": 0.06548048881813884, "chaser_new_fields": 94.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 55.60856246948242, "collisions_per_episode": 0.949999988079071, "entropy": 0.0018750086892396212, "explorer_grad_norm": 0.3297066241502762, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 122.14292907714844, "final_distance": 3.239708423614502, "max_param_abs": 2.250119686126709, "policy_loss": -6.218791421552794e-05, "value_loss": 0.07839360274374485}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:20.277443+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018340.2774565, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.476195965257195, "elapsed_seconds": 16114.305474925903, "env_steps_per_second": 4095.739658323902, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9494961138424354, "eta_seconds": 3418.1860098327675, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006283438149464853, "chaser_grad_norm": 0.019034171476960182, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 55.20565414428711, "collisions_per_episode": 1.0500000715255737, "entropy": 0.001432027667760849, "explorer_grad_norm": 0.4446190148591995, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 129.58328247070312, "final_distance": 3.658008575439453, "max_param_abs": 2.2777516841888428, "policy_loss": -0.0006528770187870236, "value_loss": 0.11454243957996368}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:26.484220+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018826.4842336, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.611858385852765, "elapsed_seconds": 16602.690189069952, "env_steps_per_second": 4095.721791205044, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8138573622093115, "eta_seconds": 2929.8865039535212, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00019735611764301695, "chaser_grad_norm": 0.014091947115957737, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.819999694824219, "chaser_subspace_norm": 56.74790954589844, "collisions_per_episode": 1.024999976158142, "entropy": 0.0021346345893107355, "explorer_grad_norm": 0.311550859361887, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 126.95128631591797, "final_distance": 3.3504257202148438, "max_param_abs": 2.2911319732666016, "policy_loss": -0.0001434955659351811, "value_loss": 0.07897400483489037}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:34.868934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019314.8689485, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.748631615085275, "elapsed_seconds": 17095.07381430699, "env_steps_per_second": 4094.7468703771547, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6783759450121821, "eta_seconds": 2442.1534020438557, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 4.781807160392226e-05, "chaser_grad_norm": 0.035776169039309025, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 57.92525100708008, "collisions_per_episode": 0.925000011920929, "entropy": 0.0018522321479395032, "explorer_grad_norm": 0.1872963383793831, "explorer_new_fields": 97.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 127.0145263671875, "final_distance": 3.466792345046997, "max_param_abs": 2.297224760055542, "policy_loss": -9.709990797279744e-05, "value_loss": 0.06055283918976784}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:47.252558+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019807.25257, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.885716232379426, "elapsed_seconds": 17588.578436565935, "env_steps_per_second": 4093.5656204207467, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428573591532696, "eta_seconds": 1954.2864929517707, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0002409969932853917, "chaser_grad_norm": 0.0705314502120018, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 59.897953033447266, "collisions_per_episode": 1.100000023841858, "entropy": 0.002494490414392203, "explorer_grad_norm": 0.20680732280015945, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 129.95164489746094, "final_distance": 3.225666046142578, "max_param_abs": 2.3122336864471436, "policy_loss": -0.00024391081492591837, "value_loss": 0.1011340543627739}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:00.757180+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020300.757196, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.022272800527475, "elapsed_seconds": 18080.182081898907, "env_steps_per_second": 4092.879134999729, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40721130815087625, "eta_seconds": 1465.9607093431546, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0006484899224137308, "chaser_grad_norm": 0.08261850755661726, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 59.100555419921875, "collisions_per_episode": 0.824999988079071, "entropy": 0.003048810991458595, "explorer_grad_norm": 0.3042645677924156, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 126.07718658447266, "final_distance": 3.82427978515625, "max_param_abs": 2.328068971633911, "policy_loss": -0.0004626124133366716, "value_loss": 0.0969211794435978}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:12.360826+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020792.3608377, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.16034138592416, "elapsed_seconds": 18577.228989326977, "env_steps_per_second": 4091.029940130665, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27159691504864003, "eta_seconds": 977.748894175104, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005912727110923451, "chaser_grad_norm": 0.03518475824967027, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 59.17195129394531, "collisions_per_episode": 1.25, "entropy": 0.0032758425804786384, "explorer_grad_norm": 0.9495732337236404, "explorer_new_fields": 97.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.67500305175781, "explorer_subspace_norm": 130.27586364746094, "final_distance": 3.5043532848358154, "max_param_abs": 2.353240966796875, "policy_loss": -0.0007552359548861354, "value_loss": 0.12067781016230583}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:29.407734+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021289.407749, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.297516541018591, "elapsed_seconds": 19071.059547666926, "env_steps_per_second": 4089.9667795092273, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13583375746201515, "eta_seconds": 489.0015268632545, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0012391397622986688, "chaser_grad_norm": 0.11135800927877426, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 53.68886184692383, "collisions_per_episode": 1.1750000715255737, "entropy": 0.003638271940872073, "explorer_grad_norm": 0.47454552724957466, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 126.65333557128906, "final_distance": 3.1356143951416016, "max_param_abs": 2.3651559352874756, "policy_loss": -0.0005628480403672143, "value_loss": 0.08898641727864742}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:43.238292+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021783.2383027, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.43102195080166, "elapsed_seconds": 19551.679022885975, "env_steps_per_second": 4091.7201999049284, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:43.857767+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022263.857778, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.431030371984137, "elapsed_seconds": 19551.709339142893, "env_steps_per_second": 4091.713855414088, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:43.888084+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022263.8880963, "update": 20000}

non_social/seed_0002/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.431030371984137,
+  "elapsed_seconds": 19551.709339142893,
+  "env_steps_per_second": 4091.713855414088,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 3.0426282704221208e-05,
+    "chaser_grad_norm": 0.03582112770527601,
+    "chaser_new_fields": 94.5250015258789,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.820000648498535,
+    "chaser_subspace_norm": 58.25376892089844,
+    "collisions_per_episode": 0.9750000238418579,
+    "entropy": 0.0015975794522091746,
+    "explorer_grad_norm": 0.33543068170547485,
+    "explorer_new_fields": 97.5250015258789,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 95.57500457763672,
+    "explorer_subspace_norm": 127.81230163574219,
+    "final_distance": 3.7922744750976562,
+    "max_param_abs": 2.384674310684204,
+    "policy_loss": -4.3445531147767724e-05,
+    "value_loss": 0.07627437077462673
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-02T19:57:43.888084+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783022263.8880963,
+  "update": 20000
+}

non_social/seed_0003/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a06bb3351d710015bd4527f14150baa0ed7be3ea4e957deb0bac2ad812840a89
+size 29760352

non_social/seed_0003/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01",
+  "seed": 3,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0003/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0006961167227644991, "elapsed_seconds": 2.5060202019521967, "env_steps_per_second": 1596.1563266265728, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.921638338567217, "eta_seconds": 50117.89801884198, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005024423047657933, "chaser_grad_norm": 5.189185708761215, "chaser_new_fields": 31.17500114440918, "chaser_partner_vision": 0.0, "chaser_return": -3.8100006580352783, "chaser_subspace_norm": 0.029627107083797455, "collisions_per_episode": 1.725000023841858, "entropy": 2.755772113800049, "explorer_grad_norm": 9.743041276931763, "explorer_new_fields": 31.0, "explorer_partner_vision": 0.0, "explorer_return": -3.799999952316284, "explorer_subspace_norm": 0.038101453334093094, "final_distance": 4.984309673309326, "max_param_abs": 1.0012016296386719, "policy_loss": -0.01039676697925529, "value_loss": 7.190685272216797}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.674447+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.6744661, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.1340119316072216, "elapsed_seconds": 482.4429537859978, "env_steps_per_second": 4145.567852747955, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.226465332681643, "eta_seconds": 18815.275197653915, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.003495984075223113, "chaser_grad_norm": 0.4796079024672508, "chaser_new_fields": 89.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.780001163482666, "chaser_subspace_norm": 11.31277084350586, "collisions_per_episode": 1.649999976158142, "entropy": 0.12384101748466492, "explorer_grad_norm": 7.731855094432831, "explorer_new_fields": 88.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 81.8125, "explorer_subspace_norm": 14.753433227539062, "final_distance": 2.7396857738494873, "max_param_abs": 1.3429911136627197, "policy_loss": -0.001576730081808364, "value_loss": 0.7917164713144302}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:39:54.611379+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003194.611392, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.26942642090693375, "elapsed_seconds": 969.9351152649615, "env_steps_per_second": 4123.987199811095, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.119101997231741, "eta_seconds": 18428.76719003427, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006954645487520228, "chaser_grad_norm": 0.3521268740296364, "chaser_new_fields": 89.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 7.770000457763672, "chaser_subspace_norm": 15.745429039001465, "collisions_per_episode": 3.625, "entropy": 0.03980019688606262, "explorer_grad_norm": 6.109767258167267, "explorer_new_fields": 90.25, "explorer_partner_vision": 0.0, "explorer_return": 84.70000457763672, "explorer_subspace_norm": 28.12783432006836, "final_distance": 2.597499132156372, "max_param_abs": 1.50963294506073, "policy_loss": -0.0020415620807305856, "value_loss": 0.4672740027308464}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:02.103538+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003682.1035526, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.40329279701804305, "elapsed_seconds": 1451.854069264955, "env_steps_per_second": 4132.646749433765, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.9739444965558635, "eta_seconds": 17906.20018760111, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004746300520819702, "chaser_grad_norm": 0.3817563150078058, "chaser_new_fields": 91.625, "chaser_partner_vision": 0.0, "chaser_return": 8.24000072479248, "chaser_subspace_norm": 20.85103988647461, "collisions_per_episode": 1.1750000715255737, "entropy": 0.026081882882863283, "explorer_grad_norm": 4.811295688152313, "explorer_new_fields": 93.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 89.3125, "explorer_subspace_norm": 39.89060592651367, "final_distance": 3.027148962020874, "max_param_abs": 1.6245406866073608, "policy_loss": -0.004421283919370467, "value_loss": 0.3692459166049957}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:04.022491+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004164.022504, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.538079933566106, "elapsed_seconds": 1937.0877608379815, "env_steps_per_second": 4129.910973439433, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8427194020949536, "eta_seconds": 17433.789847541833, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0004433605326805701, "chaser_grad_norm": 0.05365113168954849, "chaser_new_fields": 92.125, "chaser_partner_vision": 0.0, "chaser_return": 8.34000015258789, "chaser_subspace_norm": 25.149837493896484, "collisions_per_episode": 0.925000011920929, "entropy": 0.014946554088965058, "explorer_grad_norm": 2.882696211338043, "explorer_new_fields": 94.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.11250305175781, "explorer_subspace_norm": 45.11609649658203, "final_distance": 2.9697210788726807, "max_param_abs": 1.7043564319610596, "policy_loss": -0.0007080689495202463, "value_loss": 0.14073776081204414}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:09.256186+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004649.2562013, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6727541947155583, "elapsed_seconds": 2421.91510097601, "env_steps_per_second": 4128.963891413902, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.709279363008908, "eta_seconds": 16953.40570683207, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0038921834673546485, "chaser_grad_norm": 0.15854962915182114, "chaser_new_fields": 91.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.225001335144043, "chaser_subspace_norm": 27.932138442993164, "collisions_per_episode": 1.25, "entropy": 0.01696440251544118, "explorer_grad_norm": 2.074641853570938, "explorer_new_fields": 94.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.73750305175781, "explorer_subspace_norm": 52.29011154174805, "final_distance": 3.3177506923675537, "max_param_abs": 1.7815338373184204, "policy_loss": -0.0045965834856218635, "value_loss": 0.1989050954580307}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:14.083525+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005134.0835383, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8094021273358117, "elapsed_seconds": 2913.847658408922, "env_steps_per_second": 4118.266088952805, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.586612054902933, "eta_seconds": 16511.803397650558, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.005292883627254774, "chaser_grad_norm": 0.08518719859421253, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.395000457763672, "chaser_subspace_norm": 30.823591232299805, "collisions_per_episode": 0.6000000238418579, "entropy": 0.011396759422495961, "explorer_grad_norm": 0.8767628371715546, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 56.86723327636719, "final_distance": 3.3559136390686035, "max_param_abs": 1.8410812616348267, "policy_loss": -0.0024168349774527087, "value_loss": 0.15405790507793427}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:26.016083+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005626.0160983, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9458058168291528, "elapsed_seconds": 3404.90094058495, "env_steps_per_second": 4111.720206930557, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.458798850766007, "eta_seconds": 16051.675862757624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0018205504827417118, "chaser_grad_norm": 0.0703201126307249, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.34000015258789, "chaser_subspace_norm": 32.318973541259766, "collisions_per_episode": 0.9750000238418579, "entropy": 0.013418101938441396, "explorer_grad_norm": 0.6616397947072983, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.7125015258789, "explorer_subspace_norm": 61.08270263671875, "final_distance": 3.17974853515625, "max_param_abs": 1.8905514478683472, "policy_loss": -0.0017446497985957876, "value_loss": 0.146530669182539}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:37.069364+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006117.0693789, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.080710129513043, "elapsed_seconds": 3890.556466246955, "env_steps_per_second": 4112.522241692197, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.322840518052172, "eta_seconds": 15562.22586498782, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0001880493554962115, "chaser_grad_norm": 0.0595186036080122, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.370001792907715, "chaser_subspace_norm": 35.88082504272461, "collisions_per_episode": 0.9000000357627869, "entropy": 0.007024818332865834, "explorer_grad_norm": 0.5938415750861168, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 65.14606475830078, "final_distance": 3.142601728439331, "max_param_abs": 1.9357455968856812, "policy_loss": -0.000688961556879697, "value_loss": 0.10439493507146835}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:42.724890+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006602.724905, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.215635584145558, "elapsed_seconds": 4376.288102924009, "env_steps_per_second": 4113.074728323607, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.187189234279145, "eta_seconds": 15073.88124340492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0002825130148482735, "chaser_grad_norm": 0.07820950355380774, "chaser_new_fields": 92.125, "chaser_partner_vision": 0.0, "chaser_return": 8.340001106262207, "chaser_subspace_norm": 39.256492614746094, "collisions_per_episode": 1.0500000715255737, "entropy": 0.006176907219924033, "explorer_grad_norm": 1.275507390499115, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.3375015258789, "explorer_subspace_norm": 66.62557220458984, "final_distance": 3.2689292430877686, "max_param_abs": 1.9760191440582275, "policy_loss": -0.0004644948476926203, "value_loss": 0.1531773805618286}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:44:48.456525+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007088.4565394, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.349486972211388, "elapsed_seconds": 4858.153099960997, "env_steps_per_second": 4116.790802694252, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.048460916634164, "eta_seconds": 14574.459299882988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.006399905010243678, "chaser_grad_norm": 0.11409256421029568, "chaser_new_fields": 92.0, "chaser_partner_vision": 0.0, "chaser_return": 8.335000991821289, "chaser_subspace_norm": 39.89152908325195, "collisions_per_episode": 1.1750000715255737, "entropy": 0.011616653995588422, "explorer_grad_norm": 1.2792066633701324, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.26250457763672, "explorer_subspace_norm": 65.928955078125, "final_distance": 3.333908796310425, "max_param_abs": 2.0132949352264404, "policy_loss": -0.0027241093716590115, "value_loss": 0.21737461909651756}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:52:50.321523+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007570.3215399, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.4845045746044423, "elapsed_seconds": 5344.216468575993, "env_steps_per_second": 4116.599716602062, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9136938785026207, "eta_seconds": 14089.297962609435, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0012355989805410417, "chaser_grad_norm": 0.28674943186342716, "chaser_new_fields": 92.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.535000801086426, "chaser_subspace_norm": 41.7598762512207, "collisions_per_episode": 0.8500000238418579, "entropy": 0.006215645698830485, "explorer_grad_norm": 0.3715188130736351, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 69.58002471923828, "final_distance": 3.4541115760803223, "max_param_abs": 2.048832654953003, "policy_loss": -0.0007717605554518059, "value_loss": 0.09684472344815731}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:00:56.384891+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008056.3849046, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.6182306944688833, "elapsed_seconds": 5825.63050008798, "env_steps_per_second": 4119.725753227491, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7758716204273948, "eta_seconds": 13593.137833538622, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 4.920778848616081e-05, "chaser_grad_norm": 0.049862075597047806, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.445000648498535, "chaser_subspace_norm": 41.50165939331055, "collisions_per_episode": 1.024999976158142, "entropy": 0.007846252177841961, "explorer_grad_norm": 1.0918368846178055, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 76.08346557617188, "final_distance": 2.907108783721924, "max_param_abs": 2.0842835903167725, "policy_loss": -0.00036661602370813284, "value_loss": 0.1382632665336132}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:08:57.798923+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008537.7989385, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7589642375672054, "elapsed_seconds": 6332.27125524194, "env_steps_per_second": 4105.951711793276, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.653233416485734, "eta_seconds": 13151.640299348643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.002392652011237839, "chaser_grad_norm": 0.20674092881381512, "chaser_new_fields": 92.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.545001029968262, "chaser_subspace_norm": 42.17708969116211, "collisions_per_episode": 0.5, "entropy": 0.004808624275028706, "explorer_grad_norm": 0.7917798906564713, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 80.67359161376953, "final_distance": 3.240365743637085, "max_param_abs": 2.118196487426758, "policy_loss": -0.00039862397231615887, "value_loss": 0.0782720111310482}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:24.439677+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009044.4396899, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.892884754610261, "elapsed_seconds": 6814.385116596939, "env_steps_per_second": 4108.95473632741, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.515357401419056, "eta_seconds": 12655.286645108601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0031535500349347823, "chaser_grad_norm": 0.13878342136740685, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.505000114440918, "chaser_subspace_norm": 44.58604431152344, "collisions_per_episode": 0.925000011920929, "entropy": 0.013588826172053814, "explorer_grad_norm": 0.4408244490623474, "explorer_new_fields": 94.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 90.55000305175781, "explorer_subspace_norm": 78.97016906738281, "final_distance": 3.387521505355835, "max_param_abs": 2.1496760845184326, "policy_loss": -0.005201074440968512, "value_loss": 0.24234837293624878}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:26.553540+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009526.553556, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.0274851760958295, "elapsed_seconds": 7298.946633944986, "env_steps_per_second": 4110.182126894849, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3791419601597155, "eta_seconds": 12164.911056574976, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00019590932616764256, "chaser_grad_norm": 0.03927200613543391, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.555000305175781, "chaser_subspace_norm": 43.92483139038086, "collisions_per_episode": 0.9000000357627869, "entropy": 0.00862066401168704, "explorer_grad_norm": 0.3418860509991646, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 84.3548812866211, "final_distance": 3.2050235271453857, "max_param_abs": 2.185359477996826, "policy_loss": -0.0005563736307649947, "value_loss": 0.10596890561282635}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:31.115057+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010011.115069, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.161654925274165, "elapsed_seconds": 7781.957730986993, "env_steps_per_second": 4112.0757920052865, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.242482387911247, "eta_seconds": 11672.93659648049, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.008034418572132385, "chaser_grad_norm": 0.10141916945576668, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 45.13732147216797, "collisions_per_episode": 1.024999976158142, "entropy": 0.007907140767201781, "explorer_grad_norm": 0.6698158085346222, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 83.3248519897461, "final_distance": 3.2458255290985107, "max_param_abs": 2.2183961868286133, "policy_loss": -0.0011762413340576927, "value_loss": 0.12257060222327709}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:41:34.126155+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010494.1261694, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.3011383428327585, "elapsed_seconds": 8284.09803419793, "env_steps_per_second": 4104.248870503848, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1133048167737316, "eta_seconds": 11207.897340385434, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 5.7506093533055136e-05, "chaser_grad_norm": 0.06035558506846428, "chaser_new_fields": 93.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.660000801086426, "chaser_subspace_norm": 48.30387496948242, "collisions_per_episode": 0.625, "entropy": 0.007350853877142072, "explorer_grad_norm": 1.0471926033496857, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 82.67190551757812, "final_distance": 3.156557083129883, "max_param_abs": 2.2406225204467773, "policy_loss": -0.00033004977623951604, "value_loss": 0.07093416340649128}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:49:56.266458+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010996.266472, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.4384581666619245, "elapsed_seconds": 8778.449399982928, "env_steps_per_second": 4100.952042859643, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9803377592534632, "eta_seconds": 10729.215933312467, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0018212820362997206, "chaser_grad_norm": 0.08922404050827026, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 47.32369613647461, "collisions_per_episode": 0.925000011920929, "entropy": 0.00711581576615572, "explorer_grad_norm": 0.5106174945831299, "explorer_new_fields": 95.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.9124984741211, "explorer_subspace_norm": 87.91162109375, "final_distance": 2.5587334632873535, "max_param_abs": 2.281062602996826, "policy_loss": -0.0012821031226906143, "value_loss": 0.1541689820587635}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:10.617825+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011490.6178405, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.5730465846694357, "elapsed_seconds": 9262.96770480997, "env_steps_per_second": 4102.356956320574, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8438935935820084, "eta_seconds": 10238.01693689523, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0018698370508686696, "chaser_grad_norm": 0.04480297304689884, "chaser_new_fields": 93.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 48.48903274536133, "collisions_per_episode": 0.8500000238418579, "entropy": 0.003263553255237639, "explorer_grad_norm": 0.9020937532186508, "explorer_new_fields": 95.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 88.82749938964844, "final_distance": 2.9119434356689453, "max_param_abs": 2.3082261085510254, "policy_loss": -0.0006277953559781224, "value_loss": 0.06033980753272772}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:15.136128+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011975.1361413, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.708521510595003, "elapsed_seconds": 9750.677438142011, "env_steps_per_second": 4102.279072788402, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.708521510595003, "eta_seconds": 9750.677438142011, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006473027324140901, "chaser_grad_norm": 0.10710760951042175, "chaser_new_fields": 93.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.55000114440918, "chaser_subspace_norm": 47.73188781738281, "collisions_per_episode": 1.100000023841858, "entropy": 0.006507194717414677, "explorer_grad_norm": 0.37044718861579895, "explorer_new_fields": 95.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 89.69586944580078, "final_distance": 2.9019782543182373, "max_param_abs": 2.330742120742798, "policy_loss": -0.0003978462544580452, "value_loss": 0.10910336673259735}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:22.845862+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012462.845876, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.8425074297510826, "elapsed_seconds": 10233.026747103897, "env_steps_per_second": 4104.357492458098, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5717924364414557, "eta_seconds": 9258.45277118924, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005266433975092966, "chaser_grad_norm": 0.06456576567143202, "chaser_new_fields": 93.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 51.4437255859375, "collisions_per_episode": 1.024999976158142, "entropy": 0.007705657742917538, "explorer_grad_norm": 0.4191744402050972, "explorer_new_fields": 95.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 90.49082946777344, "final_distance": 3.1663146018981934, "max_param_abs": 2.3666255474090576, "policy_loss": -0.0008286951470646287, "value_loss": 0.10708197765052319}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:22:25.195170+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012945.195182, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.97652851734139, "elapsed_seconds": 10715.502662429004, "env_steps_per_second": 4106.200276938387, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.43534151418841, "eta_seconds": 8767.229451078276, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0008063959727191428, "chaser_grad_norm": 0.045782639645040035, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.565000534057617, "chaser_subspace_norm": 51.45419692993164, "collisions_per_episode": 1.25, "entropy": 0.004481318756006658, "explorer_grad_norm": 0.3449392169713974, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 90.58958435058594, "final_distance": 3.308798313140869, "max_param_abs": 2.3922159671783447, "policy_loss": -0.0008177013732479921, "value_loss": 0.15009310096502304}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:30:27.671086+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013427.6711, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.111051448044988, "elapsed_seconds": 11199.785212961957, "env_steps_per_second": 4107.221623032768, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.299472809424556, "eta_seconds": 8278.102113928402, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0007127156194903783, "chaser_grad_norm": 0.03489385172724724, "chaser_new_fields": 93.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.565000534057617, "chaser_subspace_norm": 52.208133697509766, "collisions_per_episode": 1.024999976158142, "entropy": 0.0037381406873464584, "explorer_grad_norm": 0.4758947938680649, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 99.85450744628906, "final_distance": 2.635690212249756, "max_param_abs": 2.4100501537323, "policy_loss": -0.0008028766119352326, "value_loss": 0.11823840066790581}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:38:31.953636+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013911.9536479, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.245491730141641, "elapsed_seconds": 11683.770228509908, "env_steps_per_second": 4108.262920377688, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.163661153427761, "eta_seconds": 7789.180152339939, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0020081153855908696, "chaser_grad_norm": 0.04150939965620637, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.570000648498535, "chaser_subspace_norm": 53.23170852661133, "collisions_per_episode": 0.875, "entropy": 0.004793138708919287, "explorer_grad_norm": 0.7487403303384781, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 96.56257629394531, "final_distance": 3.115834951400757, "max_param_abs": 2.4359054565429688, "policy_loss": -0.000987323350628344, "value_loss": 0.08897869288921356}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:46:35.938653+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014395.9386683, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.3880492816511025, "elapsed_seconds": 12196.977413943969, "env_steps_per_second": 4099.376288328486, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0328295689906613, "eta_seconds": 7318.18644836638, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005557573119234771, "chaser_grad_norm": 0.04495162144303322, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 55.89772415161133, "collisions_per_episode": 1.024999976158142, "entropy": 0.004529337864369154, "explorer_grad_norm": 0.43399157002568245, "explorer_new_fields": 95.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 94.74485778808594, "final_distance": 3.02925705909729, "max_param_abs": 2.445003032684326, "policy_loss": -0.00028155376063576654, "value_loss": 0.11668848805129528}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:09.145836+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014909.1458454, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5251567873161145, "elapsed_seconds": 12690.564434338012, "env_steps_per_second": 4097.5324832124, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8981613470163694, "eta_seconds": 6833.38084925893, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.004025487210471246, "chaser_grad_norm": 0.11131552793085575, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 52.96380615234375, "collisions_per_episode": 1.0500000715255737, "entropy": 0.005851391702890396, "explorer_grad_norm": 0.4314912408590317, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 93.7562484741211, "final_distance": 2.8668484687805176, "max_param_abs": 2.472851276397705, "policy_loss": -0.0010405651386875547, "value_loss": 0.12776139751076698}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:22.732860+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015402.7328765, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.662232038019445, "elapsed_seconds": 13184.035336870002, "env_steps_per_second": 4095.8628083304307, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7632969071945475, "eta_seconds": 6347.868865900371, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00040896350521180125, "chaser_grad_norm": 0.05079989228397608, "chaser_new_fields": 93.625, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 57.54914093017578, "collisions_per_episode": 1.125, "entropy": 0.005252829403616488, "explorer_grad_norm": 0.49959227815270424, "explorer_new_fields": 95.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.36250305175781, "explorer_subspace_norm": 98.56642150878906, "final_distance": 3.139416217803955, "max_param_abs": 2.5008347034454346, "policy_loss": -0.0006854486721579178, "value_loss": 0.1017832774668932}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:36.203761+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015896.203774, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.798527904025266, "elapsed_seconds": 13674.700454490958, "env_steps_per_second": 4095.153688109405, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6279405302965426, "eta_seconds": 5860.585909067553, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0022012519165470756, "chaser_grad_norm": 0.05405232170596719, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 59.76309585571289, "collisions_per_episode": 1.149999976158142, "entropy": 0.00722871208563447, "explorer_grad_norm": 0.6766604036092758, "explorer_new_fields": 95.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 104.74185943603516, "final_distance": 2.8553645610809326, "max_param_abs": 2.524658679962158, "policy_loss": -0.00035590077948199905, "value_loss": 0.12009123712778091}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:19:46.868878+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016386.8688936, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.937308964141929, "elapsed_seconds": 14174.312270910945, "env_steps_per_second": 4091.9092857175, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4934620208814213, "eta_seconds": 5376.463275173117, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004644796858217504, "chaser_grad_norm": 0.4359093941748142, "chaser_new_fields": 93.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.670001029968262, "chaser_subspace_norm": 60.4807243347168, "collisions_per_episode": 0.9750000238418579, "entropy": 0.007179249660111964, "explorer_grad_norm": 0.3641759119927883, "explorer_new_fields": 95.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0625, "explorer_subspace_norm": 100.59651947021484, "final_distance": 2.7122199535369873, "max_param_abs": 2.5459628105163574, "policy_loss": -0.0011473420108529808, "value_loss": 0.12315200828015804}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:06.480695+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016886.480711, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.074349657038596, "elapsed_seconds": 14667.658765338943, "env_steps_per_second": 4090.6323878890366, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3581165523461984, "eta_seconds": 4889.219588446314, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003340514101120573, "chaser_grad_norm": 0.07047029910609126, "chaser_new_fields": 93.75, "chaser_partner_vision": 0.0, "chaser_return": 8.695000648498535, "chaser_subspace_norm": 61.55753707885742, "collisions_per_episode": 0.925000011920929, "entropy": 0.0027071837102994323, "explorer_grad_norm": 0.37893088161945343, "explorer_new_fields": 95.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 107.49569702148438, "final_distance": 2.6742610931396484, "max_param_abs": 2.5662577152252197, "policy_loss": -0.00021887112181873114, "value_loss": 0.12173754349350929}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:19.827188+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017379.8272038, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.209102502566368, "elapsed_seconds": 15152.769009238924, "env_steps_per_second": 4091.661396157854, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2219975007450745, "eta_seconds": 4399.191002682268, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005762351906208885, "chaser_grad_norm": 0.07829482853412628, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 67.62864685058594, "collisions_per_episode": 0.875, "entropy": 0.004984446335583925, "explorer_grad_norm": 0.22619431093335152, "explorer_new_fields": 95.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 106.18888092041016, "final_distance": 2.723506450653076, "max_param_abs": 2.5889992713928223, "policy_loss": -0.00048407367150815617, "value_loss": 0.07387478835880756}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:24.937431+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017864.9374437, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.343068402223031, "elapsed_seconds": 15635.04624800291, "env_steps_per_second": 4093.3681285512557, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0857671005557576, "eta_seconds": 3908.7615620007273, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005539466711863561, "chaser_grad_norm": 0.06942280102521181, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 62.85855484008789, "collisions_per_episode": 0.875, "entropy": 0.005629157414659858, "explorer_grad_norm": 0.31974368914961815, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 110.23323822021484, "final_distance": 3.025721311569214, "max_param_abs": 2.6085193157196045, "policy_loss": -0.0007279603790912859, "value_loss": 0.1100351344794035}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:27.214672+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018347.2146873, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.476395920648033, "elapsed_seconds": 16115.025314332917, "env_steps_per_second": 4095.5567064048437, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9495385286223099, "eta_seconds": 3418.3387030403155, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0003900722442824467, "chaser_grad_norm": 0.13584143295884132, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 66.25489807128906, "collisions_per_episode": 1.0, "entropy": 0.005487023736350238, "explorer_grad_norm": 0.4194389581680298, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 110.97061920166016, "final_distance": 2.7326693534851074, "max_param_abs": 2.6253645420074463, "policy_loss": -0.000591823476483011, "value_loss": 0.12209559418261051}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:27.193738+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018827.1937509, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.612801431243586, "elapsed_seconds": 16606.08515247691, "env_steps_per_second": 4094.8844580540613, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8140237819841623, "eta_seconds": 2930.485615142984, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0004943212216756976, "chaser_grad_norm": 0.04927615146152675, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.705000877380371, "chaser_subspace_norm": 66.02193450927734, "collisions_per_episode": 0.7250000238418579, "entropy": 0.004525314900092781, "explorer_grad_norm": 0.6757950857281685, "explorer_new_fields": 95.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 112.50408172607422, "final_distance": 2.9537694454193115, "max_param_abs": 2.631089925765991, "policy_loss": -0.0002603902962299287, "value_loss": 0.06927836500108242}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:38.253576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019318.2535892, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.748318701824141, "elapsed_seconds": 17093.947326566908, "env_steps_per_second": 4095.016713384162, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6783312431177344, "eta_seconds": 2441.992475223844, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0011644642110459902, "chaser_grad_norm": 0.03152832528576255, "chaser_new_fields": 93.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.574999809265137, "chaser_subspace_norm": 71.01042175292969, "collisions_per_episode": 1.1750000715255737, "entropy": 0.005007880623452365, "explorer_grad_norm": 0.4314521551132202, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 111.36642456054688, "final_distance": 3.0410330295562744, "max_param_abs": 2.6621429920196533, "policy_loss": -0.0012926242689133005, "value_loss": 0.14316125959157944}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:46.115748+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019806.1157608, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.8837019592194375, "elapsed_seconds": 17581.327053189976, "env_steps_per_second": 4095.2540034192834, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5426335510243819, "eta_seconds": 1953.480783687775, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005835298663985355, "chaser_grad_norm": 0.07682760432362556, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.699999809265137, "chaser_subspace_norm": 75.85527801513672, "collisions_per_episode": 0.925000011920929, "entropy": 0.0035212296061217785, "explorer_grad_norm": 0.4473385736346245, "explorer_new_fields": 95.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.25, "explorer_subspace_norm": 113.15235900878906, "final_distance": 2.944685697555542, "max_param_abs": 2.6787211894989014, "policy_loss": -0.0009910173625691243, "value_loss": 0.08652106299996376}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:24:53.495476+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020293.495492, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.018611647756382, "elapsed_seconds": 18067.001931922976, "env_steps_per_second": 4095.864951962384, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40691445792619313, "eta_seconds": 1464.8920485342953, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 4.9106388614772456e-05, "chaser_grad_norm": 0.03466647490859032, "chaser_new_fields": 93.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.65999984741211, "chaser_subspace_norm": 75.69221496582031, "collisions_per_episode": 0.949999988079071, "entropy": 0.003670561476610601, "explorer_grad_norm": 0.2681044861674309, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.2874984741211, "explorer_subspace_norm": 118.05979919433594, "final_distance": 2.848959445953369, "max_param_abs": 2.7065818309783936, "policy_loss": -0.00013019449774542835, "value_loss": 0.11256165988743305}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:32:59.170354+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020779.1703684, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.152573043844152, "elapsed_seconds": 18549.262957838946, "env_steps_per_second": 4097.197833290853, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2711880549391658, "eta_seconds": 976.276997780997, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 7.057428441470392e-05, "chaser_grad_norm": 0.03550256649032235, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 77.3493423461914, "collisions_per_episode": 1.0750000476837158, "entropy": 0.003740044543519616, "explorer_grad_norm": 0.3448968231678009, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.7249984741211, "explorer_subspace_norm": 115.06394958496094, "final_distance": 2.810687780380249, "max_param_abs": 2.711897850036621, "policy_loss": -0.00033303190628153345, "value_loss": 0.12280823849141598}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:01.431381+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021261.4313953, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.289410399953582, "elapsed_seconds": 19041.877439832897, "env_steps_per_second": 4096.234746098886, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1356259076911175, "eta_seconds": 488.25326768802296, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.001099711209915366, "chaser_grad_norm": 0.020938430912792683, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.660000801086426, "chaser_subspace_norm": 81.31414794921875, "collisions_per_episode": 1.0, "entropy": 0.00502233556471765, "explorer_grad_norm": 0.39622850716114044, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 119.11949157714844, "final_distance": 3.1853957176208496, "max_param_abs": 2.7196757793426514, "policy_loss": -0.0005644467060932357, "value_loss": 0.12597087770700455}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:14.045863+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021754.0458794, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.426759173181652, "elapsed_seconds": 19536.333023453946, "env_steps_per_second": 4094.9342900716133, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:28.501446+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022248.5014582, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.427600972299988, "elapsed_seconds": 19539.363500279956, "env_steps_per_second": 4094.2991822048743, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:31.531925+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022251.531941, "update": 20000}

non_social/seed_0003/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.427600972299988,
+  "elapsed_seconds": 19539.363500279956,
+  "env_steps_per_second": 4094.2991822048743,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 0.0002613905893603663,
+    "chaser_grad_norm": 0.02223971695639193,
+    "chaser_new_fields": 93.7249984741211,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.6850004196167,
+    "chaser_subspace_norm": 82.19728088378906,
+    "collisions_per_episode": 1.0,
+    "entropy": 0.0022482319036498666,
+    "explorer_grad_norm": 0.44222569093108177,
+    "explorer_new_fields": 95.7750015258789,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 93.0999984741211,
+    "explorer_subspace_norm": 121.7215576171875,
+    "final_distance": 2.864136219024658,
+    "max_param_abs": 2.740300178527832,
+    "policy_loss": -0.00019148159428716838,
+    "value_loss": 0.08482817187905312
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-02T19:57:31.531925+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783022251.531941,
+  "update": 20000
+}

non_social/seed_0004/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fc0051ad27e22e8e395faf9f846864d859748404592f14a524f3596f1f93559
+size 29760360

non_social/seed_0004/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01",
+  "seed": 4,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0004/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0007179040744409172, "elapsed_seconds": 2.584454667987302, "env_steps_per_second": 1547.7152876955213, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.357363584743903, "eta_seconds": 51686.50890507805, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004168721541645493, "chaser_grad_norm": 6.509516932070255, "chaser_new_fields": 34.95000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.06000018119812, "chaser_subspace_norm": 0.032386090606451035, "collisions_per_episode": 2.049999952316284, "entropy": 2.7535924315452576, "explorer_grad_norm": 6.642140626907349, "explorer_new_fields": 31.42500114440918, "explorer_partner_vision": 0.0, "explorer_return": -3.3125, "explorer_subspace_norm": 0.03992241621017456, "final_distance": 5.027736663818359, "max_param_abs": 1.0011932849884033, "policy_loss": -0.013245799133557057, "value_loss": 7.792621850967407}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.576904+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.576917, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.13611594677527641, "elapsed_seconds": 490.0174083909951, "env_steps_per_second": 4081.4876487085094, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.30852192423578, "eta_seconds": 19110.67892724881, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.001165815932237635, "chaser_grad_norm": 0.1750371716916561, "chaser_new_fields": 92.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.480000495910645, "chaser_subspace_norm": 13.320653915405273, "collisions_per_episode": 0.550000011920929, "entropy": 0.1037871316075325, "explorer_grad_norm": 8.205935001373291, "explorer_new_fields": 87.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 81.2874984741211, "explorer_subspace_norm": 15.509645462036133, "final_distance": 3.4141125679016113, "max_param_abs": 1.3126949071884155, "policy_loss": -0.0033158181114174567, "value_loss": 0.5582300126552582}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:02.009857+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003202.0098708, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.27003779055110905, "elapsed_seconds": 972.1360459839925, "env_steps_per_second": 4114.65043038417, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.130718020471072, "eta_seconds": 18470.584873695858, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0019922951621613727, "chaser_grad_norm": 0.31193093210458755, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 19.159109115600586, "collisions_per_episode": 0.22500000894069672, "entropy": 0.02209476660937071, "explorer_grad_norm": 4.885534763336182, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.0999984741211, "explorer_subspace_norm": 29.250194549560547, "final_distance": 4.79291296005249, "max_param_abs": 1.4873065948486328, "policy_loss": -0.00244688695413231, "value_loss": 0.11829400807619095}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:04.128494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003684.1285083, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.4079382929050027, "elapsed_seconds": 1468.5778544580098, "env_steps_per_second": 4085.585235938579, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.031238945828368, "eta_seconds": 18112.460204982122, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0036912858272062365, "chaser_grad_norm": 0.6820209696888924, "chaser_new_fields": 87.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.540000915527344, "chaser_subspace_norm": 25.168725967407227, "collisions_per_episode": 0.7750000357627869, "entropy": 0.017583402805030346, "explorer_grad_norm": 3.2422459721565247, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 89.8375015258789, "explorer_subspace_norm": 37.35585021972656, "final_distance": 3.5126266479492188, "max_param_abs": 1.6164562702178955, "policy_loss": -0.0008014342512403516, "value_loss": 0.20651323720812798}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:20.570304+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004180.5703175, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5459863309944436, "elapsed_seconds": 1965.5507915799972, "env_steps_per_second": 4070.1059643283215, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.913876978949993, "eta_seconds": 17689.957124219975, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.001175347762921497, "chaser_grad_norm": 0.24416987970471382, "chaser_new_fields": 89.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.850000858306885, "chaser_subspace_norm": 30.818607330322266, "collisions_per_episode": 0.675000011920929, "entropy": 0.013233349192887545, "explorer_grad_norm": 1.5040321201086044, "explorer_new_fields": 94.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.8499984741211, "explorer_subspace_norm": 45.67414855957031, "final_distance": 3.4709484577178955, "max_param_abs": 1.7207281589508057, "policy_loss": -0.00155099803162706, "value_loss": 0.11385388672351837}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:37.543241+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004677.5432556, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.682896484242778, "elapsed_seconds": 2458.427343274001, "env_steps_per_second": 4067.6410581581554, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.7802753896994465, "eta_seconds": 17208.991402918007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": -7.057968711263879e-05, "chaser_grad_norm": 0.068180195055902, "chaser_new_fields": 88.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.595001220703125, "chaser_subspace_norm": 33.693416595458984, "collisions_per_episode": 0.699999988079071, "entropy": 0.006714100483804941, "explorer_grad_norm": 1.265616461634636, "explorer_new_fields": 94.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.63750457763672, "explorer_subspace_norm": 54.46622848510742, "final_distance": 4.577891826629639, "max_param_abs": 1.8177306652069092, "policy_loss": -1.9406676939226486e-05, "value_loss": 0.09330222010612488}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:50.419790+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005170.4198022, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8202148621105718, "elapsed_seconds": 2952.7735035980586, "env_steps_per_second": 4063.9757791708635, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.647884218626573, "eta_seconds": 16732.383187055664, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004725548817338096, "chaser_grad_norm": 0.2967050150036812, "chaser_new_fields": 88.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.599999904632568, "chaser_subspace_norm": 37.740169525146484, "collisions_per_episode": 3.075000047683716, "entropy": 0.011035462841391563, "explorer_grad_norm": 0.8352572470903397, "explorer_new_fields": 92.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 87.88750457763672, "explorer_subspace_norm": 57.533329010009766, "final_distance": 4.167410373687744, "max_param_abs": 1.8849718570709229, "policy_loss": 0.0017718010931275785, "value_loss": 0.351640447974205}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:04.765951+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005664.765965, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9561326472975068, "elapsed_seconds": 3442.0775302710244, "env_steps_per_second": 4067.3110576035338, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.507482480116818, "eta_seconds": 16226.936928420544, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005448501374183401, "chaser_grad_norm": 0.06794039160013199, "chaser_new_fields": 91.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.305000305175781, "chaser_subspace_norm": 39.59990692138672, "collisions_per_episode": 0.699999988079071, "entropy": 0.006905361427925527, "explorer_grad_norm": 0.8802538886666298, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 65.91787719726562, "final_distance": 4.052903175354004, "max_param_abs": 1.9597896337509155, "policy_loss": -0.00037312930973776304, "value_loss": 0.10577482357621193}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:14.069979+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006154.069993, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0919844198711022, "elapsed_seconds": 3931.143911535968, "env_steps_per_second": 4070.062139686083, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.367937679484409, "eta_seconds": 15724.575646143872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007529899503896237, "chaser_grad_norm": 0.05311888735741377, "chaser_new_fields": 91.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.215001106262207, "chaser_subspace_norm": 42.80828094482422, "collisions_per_episode": 0.875, "entropy": 0.008078583749011159, "explorer_grad_norm": 0.58997692912817, "explorer_new_fields": 94.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.48750305175781, "explorer_subspace_norm": 70.58789825439453, "final_distance": 4.197776794433594, "max_param_abs": 2.0185790061950684, "policy_loss": -0.000664450679323636, "value_loss": 0.13967562094330788}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:23.136360+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006643.136374, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.2263701481272196, "elapsed_seconds": 4414.93253325799, "env_steps_per_second": 4077.07249531103, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.224163843549312, "eta_seconds": 15206.989836777522, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00033688677544142787, "chaser_grad_norm": 0.046388428658246994, "chaser_new_fields": 91.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.17500114440918, "chaser_subspace_norm": 45.70652389526367, "collisions_per_episode": 1.0, "entropy": 0.008538259658962488, "explorer_grad_norm": 0.6624040454626083, "explorer_new_fields": 94.25, "explorer_partner_vision": 0.0, "explorer_return": 91.07500457763672, "explorer_subspace_norm": 73.56023406982422, "final_distance": 4.0290141105651855, "max_param_abs": 2.0689001083374023, "policy_loss": -0.0007023077382442011, "value_loss": 0.17052556574344635}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:26.924980+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007126.9249928, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.3600353067249913, "elapsed_seconds": 4896.127104209969, "env_steps_per_second": 4084.861273883773, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.080105920174974, "eta_seconds": 14688.381312629906, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00607271511547558, "chaser_grad_norm": 0.15301267057657242, "chaser_new_fields": 91.375, "chaser_partner_vision": 0.0, "chaser_return": 8.210000991821289, "chaser_subspace_norm": 47.38425827026367, "collisions_per_episode": 0.7250000238418579, "entropy": 0.010898770997300744, "explorer_grad_norm": 0.9010711759328842, "explorer_new_fields": 94.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 90.92500305175781, "explorer_subspace_norm": 78.74980163574219, "final_distance": 3.634722948074341, "max_param_abs": 2.1259748935699463, "policy_loss": -0.0020766167314718587, "value_loss": 0.18118872493505478}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:28.119552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007608.119565, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.4964520221661142, "elapsed_seconds": 5387.227279798011, "env_steps_per_second": 4083.7334044731206, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9451916948015735, "eta_seconds": 14202.690101285665, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.001186162677462077, "chaser_grad_norm": 0.07213358487933874, "chaser_new_fields": 91.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.160000801086426, "chaser_subspace_norm": 53.68938446044922, "collisions_per_episode": 1.4250000715255737, "entropy": 0.007026613922789693, "explorer_grad_norm": 0.4948671758174896, "explorer_new_fields": 93.75, "explorer_partner_vision": 0.0, "explorer_return": 90.17500305175781, "explorer_subspace_norm": 82.62323760986328, "final_distance": 3.7338340282440186, "max_param_abs": 2.1736793518066406, "policy_loss": -0.0012927470467936075, "value_loss": 0.2303939014673233}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:39.219728+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008099.2197392, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.6302905221366544, "elapsed_seconds": 5869.045879691956, "env_steps_per_second": 4089.25070479423, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.80401121831886, "eta_seconds": 13694.440385947895, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005560482332418459, "chaser_grad_norm": 0.06472470052540302, "chaser_new_fields": 91.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 56.36345672607422, "collisions_per_episode": 0.824999988079071, "entropy": 0.006473298300988972, "explorer_grad_norm": 0.36392417550086975, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 84.27880859375, "final_distance": 3.650900363922119, "max_param_abs": 2.226118564605713, "policy_loss": -0.0016544239111198067, "value_loss": 0.14978458732366562}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:41.038327+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008581.038341, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7643612070447288, "elapsed_seconds": 6351.700345361023, "env_steps_per_second": 4093.392097596221, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6644425069390523, "eta_seconds": 13191.993024980588, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00018242453012728976, "chaser_grad_norm": 0.06973688397556543, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 55.79123306274414, "collisions_per_episode": 0.625, "entropy": 0.004662334104068577, "explorer_grad_norm": 0.2566753923892975, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 86.42034149169922, "final_distance": 3.713762044906616, "max_param_abs": 2.2646422386169434, "policy_loss": -8.36074383117058e-05, "value_loss": 0.076368048787117}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:43.692791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009063.6928043, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.8983404127383254, "elapsed_seconds": 6834.025485857972, "env_steps_per_second": 4097.1459731811, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5254893379426044, "eta_seconds": 12691.761616593376, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004217445315669366, "chaser_grad_norm": 0.26001538522541523, "chaser_new_fields": 91.75, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 55.53413391113281, "collisions_per_episode": 0.925000011920929, "entropy": 0.006662545842118561, "explorer_grad_norm": 0.5009182542562485, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 90.10924530029297, "final_distance": 3.585752248764038, "max_param_abs": 2.3055925369262695, "policy_loss": -0.0016401482354879704, "value_loss": 0.11595870740711689}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:46.017934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009546.0179505, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.034343741137224, "elapsed_seconds": 7323.637468094006, "env_steps_per_second": 4096.3251022046525, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.390572901895373, "eta_seconds": 12206.062446823344, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.001021832967690317, "chaser_grad_norm": 0.12410663813352585, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.385001182556152, "chaser_subspace_norm": 58.660423278808594, "collisions_per_episode": 0.75, "entropy": 0.0037105209194123745, "explorer_grad_norm": 0.5163974612951279, "explorer_new_fields": 94.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 94.83783721923828, "final_distance": 4.33816385269165, "max_param_abs": 2.3477609157562256, "policy_loss": -0.00017890909968798496, "value_loss": 0.12360158190131187}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:55.629915+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010035.6299274, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.170913896232507, "elapsed_seconds": 7815.2900264370255, "env_steps_per_second": 4094.537744824901, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2563708443487607, "eta_seconds": 11722.935039655538, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005332072360353379, "chaser_grad_norm": 0.07196172885596752, "chaser_new_fields": 92.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 59.72168731689453, "collisions_per_episode": 0.6000000238418579, "entropy": 0.004704791703261435, "explorer_grad_norm": 0.3538786545395851, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 100.75301361083984, "final_distance": 3.4285435676574707, "max_param_abs": 2.377171277999878, "policy_loss": -0.0005098494933299946, "value_loss": 0.083933986723423}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:07.282474+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010527.282488, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.3048166720188843, "elapsed_seconds": 8297.340019267984, "env_steps_per_second": 4097.698771057424, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1182813797902553, "eta_seconds": 11225.81296724492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0018567237811510928, "chaser_grad_norm": 0.2961665950715542, "chaser_new_fields": 92.0, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 61.537147521972656, "collisions_per_episode": 0.45000001788139343, "entropy": 0.007917658076621592, "explorer_grad_norm": 0.24600420892238617, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 102.79935455322266, "final_distance": 3.600990056991577, "max_param_abs": 2.410496950149536, "policy_loss": -0.00022455692280232142, "value_loss": 0.047318145632743835}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:09.332466+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011009.3324792, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.444835862054994, "elapsed_seconds": 8801.409103397978, "env_steps_per_second": 4090.2541373859563, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9881327202894368, "eta_seconds": 10757.277793041972, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.01082343613169795, "chaser_grad_norm": 0.07303811423480511, "chaser_new_fields": 92.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.350001335144043, "chaser_subspace_norm": 64.40133666992188, "collisions_per_episode": 1.125, "entropy": 0.005557489930652082, "explorer_grad_norm": 0.3703003525733948, "explorer_new_fields": 93.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 89.9124984741211, "explorer_subspace_norm": 101.63013458251953, "final_distance": 3.673555374145508, "max_param_abs": 2.434584379196167, "policy_loss": -0.0057121851332482265, "value_loss": 0.17832371965050697}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:33.401552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011513.4015658, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.581718608397512, "elapsed_seconds": 9294.186990231043, "env_steps_per_second": 4088.5770901684177, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8534784619130393, "eta_seconds": 10272.522462886942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.01158037428318126, "chaser_grad_norm": 0.42690640687942505, "chaser_new_fields": 89.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.835000514984131, "chaser_subspace_norm": 64.35760498046875, "collisions_per_episode": 3.1500000953674316, "entropy": 0.009429939789697528, "explorer_grad_norm": 0.4501849487423897, "explorer_new_fields": 92.75, "explorer_partner_vision": 0.0, "explorer_return": 88.9000015258789, "explorer_subspace_norm": 105.53968811035156, "final_distance": 3.7971291542053223, "max_param_abs": 2.472315549850464, "policy_loss": -0.010460890986955462, "value_loss": 0.36160628497600555}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:46.179438+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012006.1794524, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.7168194757572364, "elapsed_seconds": 9780.55011272605, "env_steps_per_second": 4089.749506825147, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7168194757572364, "eta_seconds": 9780.55011272605, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.000252617621805222, "chaser_grad_norm": 0.02687776880338788, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.51500129699707, "chaser_subspace_norm": 65.24219512939453, "collisions_per_episode": 0.4749999940395355, "entropy": 0.002895449462812394, "explorer_grad_norm": 0.38908619061112404, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 106.086669921875, "final_distance": 3.815507650375366, "max_param_abs": 2.503999948501587, "policy_loss": -2.349901598819315e-05, "value_loss": 0.0640158373862505}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:52.542562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012492.5425782, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.8534589431105672, "elapsed_seconds": 10272.452195198042, "env_steps_per_second": 4088.605057673893, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.581700948528608, "eta_seconds": 9294.12341470299, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.0002584837099098153, "chaser_grad_norm": 0.07851303089410067, "chaser_new_fields": 92.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 68.64289855957031, "collisions_per_episode": 0.5250000357627869, "entropy": 0.002932213246822357, "explorer_grad_norm": 0.16077022813260555, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 108.86304473876953, "final_distance": 3.8342368602752686, "max_param_abs": 2.519569158554077, "policy_loss": -0.0005988772140135268, "value_loss": 0.06520954519510269}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:04.444644+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012984.4446592, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.992207642904153, "elapsed_seconds": 10771.94751445495, "env_steps_per_second": 4084.683845790754, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4481698896488524, "eta_seconds": 8813.411602735869, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0028229508196702113, "chaser_grad_norm": 0.338572159409523, "chaser_new_fields": 92.625, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 65.50027465820312, "collisions_per_episode": 0.625, "entropy": 0.005788428126834333, "explorer_grad_norm": 0.46209583431482315, "explorer_new_fields": 94.875, "explorer_partner_vision": 0.0, "explorer_return": 91.9375, "explorer_subspace_norm": 112.18248748779297, "final_distance": 3.9190170764923096, "max_param_abs": 2.553093671798706, "policy_loss": -0.0014898449838938177, "value_loss": 0.08572590164840221}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:23.939964+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013483.9399781, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.1285971167527733, "elapsed_seconds": 11262.949620309984, "env_steps_per_second": 4084.187672920974, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3124413471650938, "eta_seconds": 8324.788849794337, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00043558126886455995, "chaser_grad_norm": 0.053662254475057125, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 70.17605590820312, "collisions_per_episode": 0.32500001788139343, "entropy": 0.0020849284483119845, "explorer_grad_norm": 0.3590114116668701, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 116.16581726074219, "final_distance": 4.035409450531006, "max_param_abs": 2.578110933303833, "policy_loss": -0.0006839056420631096, "value_loss": 0.02623558323830366}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:34.942069+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013974.9420834, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.267499841000568, "elapsed_seconds": 11762.999427602044, "env_steps_per_second": 4080.5918843596405, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.178333227333712, "eta_seconds": 7841.9996184013635, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0020530864643185653, "chaser_grad_norm": 0.09303068090230227, "chaser_new_fields": 92.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.375000953674316, "chaser_subspace_norm": 71.1871337890625, "collisions_per_episode": 0.675000011920929, "entropy": 0.0027941627195104957, "explorer_grad_norm": 0.37502408772706985, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 121.38587951660156, "final_distance": 3.820488691329956, "max_param_abs": 2.6030306816101074, "policy_loss": -0.0004895105308387881, "value_loss": 0.07220060378313065}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:54.991876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014474.9918914, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.401548980147765, "elapsed_seconds": 12245.576328531955, "env_steps_per_second": 4083.1071285310572, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.040929388088659, "eta_seconds": 7347.345797119174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00044117828131540193, "chaser_grad_norm": 0.07562015578150749, "chaser_new_fields": 92.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.4350004196167, "chaser_subspace_norm": 68.97872161865234, "collisions_per_episode": 0.550000011920929, "entropy": 0.002834049693774432, "explorer_grad_norm": 0.21107512712478638, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 120.04378509521484, "final_distance": 3.180603504180908, "max_param_abs": 2.619825839996338, "policy_loss": -0.00039526084764318625, "value_loss": 0.06872337684035301}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:57.568775+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014957.5687876, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5361426051116562, "elapsed_seconds": 12730.113378401962, "env_steps_per_second": 4084.8025822160957, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9040767873678148, "eta_seconds": 6854.676434524134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 4.7623881740954666e-05, "chaser_grad_norm": 0.06886796839535236, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 67.8337173461914, "collisions_per_episode": 0.45000001788139343, "entropy": 0.003157528582960367, "explorer_grad_norm": 0.4245411716401577, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 119.1957778930664, "final_distance": 3.7806153297424316, "max_param_abs": 2.649846315383911, "policy_loss": -0.00031039492675333236, "value_loss": 0.09159622341394424}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:04:02.105827+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015442.1058404, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.6700703145152915, "elapsed_seconds": 13212.25313225505, "env_steps_per_second": 4087.115154353946, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7670708921740292, "eta_seconds": 6361.455211826506, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 4.650597816358015e-05, "chaser_grad_norm": 0.06667145155370235, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.630000114440918, "chaser_subspace_norm": 71.45598602294922, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0034325195010751486, "explorer_grad_norm": 0.2968932166695595, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 119.08173370361328, "final_distance": 3.7372214794158936, "max_param_abs": 2.6703202724456787, "policy_loss": -0.00023720909302937798, "value_loss": 0.04756408557295799}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:12:04.245579+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015924.245592, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.8039523158305544, "elapsed_seconds": 13694.228336989996, "env_steps_per_second": 4089.31402499959, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6302652782130949, "eta_seconds": 5868.955001567141, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -1.3820697354560707e-05, "chaser_grad_norm": 0.04255277942866087, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.630000114440918, "chaser_subspace_norm": 72.71815490722656, "collisions_per_episode": 0.45000001788139343, "entropy": 0.0042581233428791165, "explorer_grad_norm": 0.27306586131453514, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 121.31263732910156, "final_distance": 3.9743010997772217, "max_param_abs": 2.698016405105591, "policy_loss": -0.00027522624793263617, "value_loss": 0.04279269836843014}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:06.220785+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016406.2208, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.9384114492808457, "elapsed_seconds": 14178.281217411044, "env_steps_per_second": 4090.763831710118, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4938802048996311, "eta_seconds": 5377.968737638672, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00015370788120971923, "chaser_grad_norm": 0.03278797212988138, "chaser_new_fields": 92.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 78.8346939086914, "collisions_per_episode": 0.5, "entropy": 0.004240104230120778, "explorer_grad_norm": 0.3519405126571655, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 130.72972106933594, "final_distance": 4.106181621551514, "max_param_abs": 2.7262022495269775, "policy_loss": -0.00016195154929476985, "value_loss": 0.043409342877566814}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:10.273665+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016890.2736769, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.075734397235566, "elapsed_seconds": 14672.643830048037, "env_steps_per_second": 4089.242586065252, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3585781324118553, "eta_seconds": 4890.8812766826795, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 1.1531024195349238e-05, "chaser_grad_norm": 0.033809758722782135, "chaser_new_fields": 93.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 78.26759338378906, "collisions_per_episode": 0.550000011920929, "entropy": 0.0020056855864822865, "explorer_grad_norm": 0.1279930453747511, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 126.9104232788086, "final_distance": 4.412608623504639, "max_param_abs": 2.740053176879883, "policy_loss": -9.81771981753976e-05, "value_loss": 0.0635758638381958}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:24.636278+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017384.6362925, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.212531428042211, "elapsed_seconds": 15165.113140951958, "env_steps_per_second": 4088.330856732934, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2229929952380612, "eta_seconds": 4402.77478285702, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0032838389529202816, "chaser_grad_norm": 0.06836945004761219, "chaser_new_fields": 93.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.690001487731934, "chaser_subspace_norm": 80.9717025756836, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0026452502934262156, "explorer_grad_norm": 0.13407007232308388, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 127.62902069091797, "final_distance": 4.132334232330322, "max_param_abs": 2.758204698562622, "policy_loss": -0.00015382529371521514, "value_loss": 0.04261405952274799}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:37.105588+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017877.1055987, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.3478986613297215, "elapsed_seconds": 15652.435180786997, "env_steps_per_second": 4088.8206378620575, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0869746653324304, "eta_seconds": 3913.1087951967493, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0012575061750773608, "chaser_grad_norm": 0.028763120993971825, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 80.18560028076172, "collisions_per_episode": 0.45000001788139343, "entropy": 0.0024889467167668045, "explorer_grad_norm": 0.20823930203914642, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 136.8090362548828, "final_distance": 3.9617085456848145, "max_param_abs": 2.774733066558838, "policy_loss": -0.0006444373294982597, "value_loss": 0.030278190039098263}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:44.427629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018364.4276435, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.481625659188624, "elapsed_seconds": 16133.852373079048, "env_steps_per_second": 4090.777482886085, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9506478671006173, "eta_seconds": 3422.332321562222, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -9.181823818860124e-05, "chaser_grad_norm": 0.06706667691469193, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.644999504089355, "chaser_subspace_norm": 82.16859436035156, "collisions_per_episode": 0.5, "entropy": 0.0034693812485784292, "explorer_grad_norm": 0.3202550671994686, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 135.02574157714844, "final_distance": 4.19083309173584, "max_param_abs": 2.788562297821045, "policy_loss": -0.0003576974979360159, "value_loss": 0.04477046988904476}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:45.844820+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018845.844832, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.618225405726407, "elapsed_seconds": 16625.611460615066, "env_steps_per_second": 4090.075132640224, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.814980953951719, "eta_seconds": 2933.9314342261882, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0007968248343219742, "chaser_grad_norm": 0.10342094302177429, "chaser_new_fields": 93.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 84.98808288574219, "collisions_per_episode": 0.125, "entropy": 0.002209584228694439, "explorer_grad_norm": 0.29085009172558784, "explorer_new_fields": 95.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.4375, "explorer_subspace_norm": 134.9689483642578, "final_distance": 4.616744518280029, "max_param_abs": 2.800830602645874, "policy_loss": -0.0004560549503667133, "value_loss": 0.01913414103910327}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:57.603908+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019337.603923, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.7545037821358225, "elapsed_seconds": 17116.21361568896, "env_steps_per_second": 4089.6895523573635, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6792148260194032, "eta_seconds": 2445.1733736698516, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 8.018034570356214e-05, "chaser_grad_norm": 0.047528491355478764, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 83.74442291259766, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0029699402512051165, "explorer_grad_norm": 0.2107248269021511, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 131.8700714111328, "final_distance": 4.045284271240234, "max_param_abs": 2.8165388107299805, "policy_loss": -0.0002612216699064973, "value_loss": 0.08247810043394566}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:17:08.206063+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019828.2060747, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.888731290616658, "elapsed_seconds": 17599.432646219968, "env_steps_per_second": 4091.0409697476393, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5431923656240731, "eta_seconds": 1955.4925162466632, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0002844389394993929, "chaser_grad_norm": 0.036604116670787334, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 85.01388549804688, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0026826930115930736, "explorer_grad_norm": 0.2773788385093212, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 128.55809020996094, "final_distance": 3.861576795578003, "max_param_abs": 2.830378532409668, "policy_loss": -0.00023015607497656188, "value_loss": 0.06609669700264931}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:11.425095+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020311.42511, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.0254162651230585, "elapsed_seconds": 18091.49855444301, "env_steps_per_second": 4090.3189847602025, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40746618365862636, "eta_seconds": 1466.8782611710549, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0001132674199098016, "chaser_grad_norm": 0.03674895199947059, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 86.63163757324219, "collisions_per_episode": 0.2750000059604645, "entropy": 0.0023306639050133526, "explorer_grad_norm": 0.16000011935830116, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.2125015258789, "explorer_subspace_norm": 133.56771850585938, "final_distance": 3.9755799770355225, "max_param_abs": 2.8445000648498535, "policy_loss": -0.00010502004870716064, "value_loss": 0.024576107505708933}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:23.491002+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020803.4910142, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.163160475746392, "elapsed_seconds": 18587.37771268701, "env_steps_per_second": 4088.7962344535235, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2717452881971785, "eta_seconds": 978.2830375098428, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 1.165527411541023e-05, "chaser_grad_norm": 0.02973249088972807, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 84.8570327758789, "collisions_per_episode": 0.17499999701976776, "entropy": 0.001998357241973281, "explorer_grad_norm": 0.322164174169302, "explorer_new_fields": 95.75, "explorer_partner_vision": 0.0, "explorer_return": 93.4749984741211, "explorer_subspace_norm": 131.4634552001953, "final_distance": 4.490052223205566, "max_param_abs": 2.842074155807495, "policy_loss": -3.848600599987151e-05, "value_loss": 0.019126664847135544}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:39.370162+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021299.370177, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.2997301635391585, "elapsed_seconds": 19079.02858874097, "env_steps_per_second": 4088.258458086793, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13589051701382457, "eta_seconds": 489.2058612497685, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0015112189614507887, "chaser_grad_norm": 0.03277226723730564, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.86500072479248, "chaser_subspace_norm": 89.1612548828125, "collisions_per_episode": 0.20000000298023224, "entropy": 0.0019229634199291468, "explorer_grad_norm": 0.13945288211107254, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 139.1415252685547, "final_distance": 4.151286602020264, "max_param_abs": 2.8724098205566406, "policy_loss": -0.0001659054823957007, "value_loss": 0.02723497338593006}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:51.021035+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021791.021046, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.433286101221391, "elapsed_seconds": 19559.82996439701, "env_steps_per_second": 4090.0151047129125, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:51.822411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022271.8224213, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.433294409401101, "elapsed_seconds": 19559.85987384396, "env_steps_per_second": 4090.0088505735375, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:51.852322+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022271.8523347, "update": 20000}

non_social/seed_0004/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.433294409401101,
+  "elapsed_seconds": 19559.85987384396,
+  "env_steps_per_second": 4090.0088505735375,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 0.0017689612679874234,
+    "chaser_grad_norm": 0.025046652648597956,
+    "chaser_new_fields": 94.0,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.77500057220459,
+    "chaser_subspace_norm": 89.04515075683594,
+    "collisions_per_episode": 0.4000000059604645,
+    "entropy": 0.0025691293994896114,
+    "explorer_grad_norm": 0.1489477176219225,
+    "explorer_new_fields": 95.05000305175781,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 92.38750457763672,
+    "explorer_subspace_norm": 140.36648559570312,
+    "final_distance": 3.9162888526916504,
+    "max_param_abs": 2.88704776763916,
+    "policy_loss": -0.0007027814439490854,
+    "value_loss": 0.05476926174014807
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-02T19:57:51.852322+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783022271.8523347,
+  "update": 20000
+}

non_social/seed_0005/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac8c8ce4aab7af2d19fc15778486415942ca2cc3a405b65f8c5b23c7cfb69376
+size 29760360

non_social/seed_0005/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01",
+  "seed": 5,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0005/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0006920439502573572, "elapsed_seconds": 2.491358220926486, "env_steps_per_second": 1605.549923090739, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.840186961196887, "eta_seconds": 49824.67306030879, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.008184439225479068, "chaser_grad_norm": 6.440325319766998, "chaser_new_fields": 34.04999923706055, "chaser_partner_vision": 0.0, "chaser_return": -3.2350003719329834, "chaser_subspace_norm": 0.033832646906375885, "collisions_per_episode": 1.225000023841858, "entropy": 2.7503525018692017, "explorer_grad_norm": 8.511807143688202, "explorer_new_fields": 34.72500228881836, "explorer_partner_vision": 0.0, "explorer_return": 1.8625000715255737, "explorer_subspace_norm": 0.039189036935567856, "final_distance": 5.033384799957275, "max_param_abs": 1.0011991262435913, "policy_loss": -0.008689469220115598, "value_loss": 7.674896478652954}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:48.329611+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022268.3296232, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.13179484498415453, "elapsed_seconds": 474.4614419429563, "env_steps_per_second": 4215.305656471989, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.139998954382026, "eta_seconds": 18503.996235775296, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0009361219235302087, "chaser_grad_norm": 0.45476044714450836, "chaser_new_fields": 91.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.145001411437988, "chaser_subspace_norm": 11.66552734375, "collisions_per_episode": 1.2000000476837158, "entropy": 0.07734057679772377, "explorer_grad_norm": 10.573444128036499, "explorer_new_fields": 92.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 88.67500305175781, "explorer_subspace_norm": 15.748729705810547, "final_distance": 2.8839144706726074, "max_param_abs": 1.2961766719818115, "policy_loss": -0.0011633181998043085, "value_loss": 0.42136064916849136}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:05:40.299696+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022740.299709, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.26813214382108225, "elapsed_seconds": 965.275717755896, "env_steps_per_second": 4143.893735666871, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.094510732600561, "eta_seconds": 18340.23863736202, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00010687408780452046, "chaser_grad_norm": 0.22473064064979553, "chaser_new_fields": 90.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.930000305175781, "chaser_subspace_norm": 15.033252716064453, "collisions_per_episode": 2.5, "entropy": 0.021731394343078136, "explorer_grad_norm": 5.488495588302612, "explorer_new_fields": 94.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.9625015258789, "explorer_subspace_norm": 30.144948959350586, "final_distance": 3.47773814201355, "max_param_abs": 1.433841586112976, "policy_loss": -0.0012641563932476174, "value_loss": 0.3978065028786659}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:13:51.113971+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023231.1139865, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.4040637830274904, "elapsed_seconds": 1454.6296188989654, "env_steps_per_second": 4124.761328963936, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.983453324005715, "eta_seconds": 17940.431966420572, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0015242458913891643, "chaser_grad_norm": 0.30585379526019096, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.315000534057617, "chaser_subspace_norm": 17.891708374023438, "collisions_per_episode": 1.9500000476837158, "entropy": 0.017732564825564623, "explorer_grad_norm": 4.706849932670593, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 39.83221435546875, "final_distance": 3.36601185798645, "max_param_abs": 1.5062170028686523, "policy_loss": -0.0018727443481338923, "value_loss": 0.41987208276987076}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:00.467872+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023720.4678838, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5393184688335815, "elapsed_seconds": 1941.5464878008934, "env_steps_per_second": 4120.426706373257, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.853866219502233, "eta_seconds": 17473.91839020804, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0005957404282430634, "chaser_grad_norm": 0.07774509489536285, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.535000801086426, "chaser_subspace_norm": 20.057270050048828, "collisions_per_episode": 1.274999976158142, "entropy": 0.009850812144577503, "explorer_grad_norm": 2.4770524203777313, "explorer_new_fields": 96.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.26250457763672, "explorer_subspace_norm": 50.30194091796875, "final_distance": 3.6079468727111816, "max_param_abs": 1.5557539463043213, "policy_loss": -0.0008870726244012417, "value_loss": 0.19083719700574875}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:07.384743+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024207.3847592, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6764874987074937, "elapsed_seconds": 2435.3549953469774, "env_steps_per_second": 4106.177546643564, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.735412490952456, "eta_seconds": 17047.48496742884, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0020468066642520544, "chaser_grad_norm": 0.18242399487644434, "chaser_new_fields": 92.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.444999694824219, "chaser_subspace_norm": 23.863012313842773, "collisions_per_episode": 1.4250000715255737, "entropy": 0.006107415771111846, "explorer_grad_norm": 2.5438318848609924, "explorer_new_fields": 96.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.11250305175781, "explorer_subspace_norm": 55.08370590209961, "final_distance": 3.9846444129943848, "max_param_abs": 1.595381498336792, "policy_loss": -0.0010705109410906744, "value_loss": 0.18556535616517067}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:21.193247+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024701.19326, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8146118185836045, "elapsed_seconds": 2932.6025469009764, "env_steps_per_second": 4091.9285201743355, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.6161336386404255, "eta_seconds": 16618.081099105533, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0012350974433408268, "chaser_grad_norm": 0.10789668932557106, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.445000648498535, "chaser_subspace_norm": 25.165456771850586, "collisions_per_episode": 1.0750000476837158, "entropy": 0.009079070296138525, "explorer_grad_norm": 1.4245552122592926, "explorer_new_fields": 97.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 61.17123031616211, "final_distance": 3.75791597366333, "max_param_abs": 1.6309267282485962, "policy_loss": -0.000544090077980286, "value_loss": 0.15173404291272163}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:38.440800+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025198.4408138, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9537632968552578, "elapsed_seconds": 3433.5478686789284, "env_steps_per_second": 4077.415121457607, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.496312685174788, "eta_seconds": 16186.725666629234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.002441858525618623, "chaser_grad_norm": 0.08779280353337526, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.555000305175781, "chaser_subspace_norm": 27.106536865234375, "collisions_per_episode": 1.350000023841858, "entropy": 0.005455998471006751, "explorer_grad_norm": 1.6851049661636353, "explorer_new_fields": 96.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.48750305175781, "explorer_subspace_norm": 64.252197265625, "final_distance": 4.024997234344482, "max_param_abs": 1.661129117012024, "policy_loss": -0.0019822102063358216, "value_loss": 0.2504864037036896}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:59.386121+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025699.3861337, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0913558711752476, "elapsed_seconds": 3928.8811362308916, "env_steps_per_second": 4072.4062258980275, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.36542348470099, "eta_seconds": 15715.524544923566, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0006985393587002342, "chaser_grad_norm": 0.044110307935625315, "chaser_new_fields": 93.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 28.84015655517578, "collisions_per_episode": 1.524999976158142, "entropy": 0.004622174659743905, "explorer_grad_norm": 2.166859894990921, "explorer_new_fields": 96.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.26250457763672, "explorer_subspace_norm": 72.42479705810547, "final_distance": 3.289487600326538, "max_param_abs": 1.6978211402893066, "policy_loss": -0.0011017804748902904, "value_loss": 0.20734801143407822}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:14.719390+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026194.7194054, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.226758298336095, "elapsed_seconds": 4416.329874009942, "env_steps_per_second": 4075.7824966676117, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.225500805379883, "eta_seconds": 15211.802899367578, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0034291903907175403, "chaser_grad_norm": 0.09173675160855055, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.520000457763672, "chaser_subspace_norm": 26.406984329223633, "collisions_per_episode": 1.2000000476837158, "entropy": 0.006392874172888696, "explorer_grad_norm": 0.6193827465176582, "explorer_new_fields": 96.875, "explorer_partner_vision": 0.0, "explorer_return": 94.5625, "explorer_subspace_norm": 76.7430191040039, "final_distance": 3.7917816638946533, "max_param_abs": 1.7314627170562744, "policy_loss": -0.0021327687255165984, "value_loss": 0.21480071172118187}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:22.168129+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026682.1681428, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.3627622604108183, "elapsed_seconds": 4905.944137478946, "env_steps_per_second": 4076.68726743341, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.088286781232455, "eta_seconds": 14717.832412436837, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0017705824037278495, "chaser_grad_norm": 0.11722816992551088, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 30.065780639648438, "collisions_per_episode": 1.5500000715255737, "entropy": 0.00480334204621613, "explorer_grad_norm": 1.15304896235466, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 77.98207092285156, "final_distance": 3.5316147804260254, "max_param_abs": 1.7599422931671143, "policy_loss": -0.00027412737708054635, "value_loss": 0.16086706891655922}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:31.782392+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027171.7824078, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.4970714932180513, "elapsed_seconds": 5389.457375584985, "env_steps_per_second": 4082.0436023231496, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.946824845756681, "eta_seconds": 14208.569444724051, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0004030766968986521, "chaser_grad_norm": 0.12016704119741917, "chaser_new_fields": 94.375, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 32.22527313232422, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007395019638352096, "explorer_grad_norm": 2.0366511791944504, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9749984741211, "explorer_subspace_norm": 82.64091491699219, "final_distance": 3.48870587348938, "max_param_abs": 1.7974824905395508, "policy_loss": -0.00045716741646373293, "value_loss": 0.17801041528582573}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:35.295629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027655.295642, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.631091633502478, "elapsed_seconds": 5871.929880608921, "env_steps_per_second": 4087.2422675304824, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.8058804781724485, "eta_seconds": 13701.169721420814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": -1.0344755381197501e-06, "chaser_grad_norm": 0.13730387203395367, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 33.81676483154297, "collisions_per_episode": 1.3250000476837158, "entropy": 0.0025091858115047216, "explorer_grad_norm": 0.5365075767040253, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 87.90255737304688, "final_distance": 3.928429365158081, "max_param_abs": 1.8272534608840942, "policy_loss": -0.00017002964968426681, "value_loss": 0.12052592821419239}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:37.768134+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028137.7681503, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7660568435352697, "elapsed_seconds": 6357.804636726971, "env_steps_per_second": 4089.4619268114106, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6679642134963295, "eta_seconds": 13204.671168586787, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.000521471640998819, "chaser_grad_norm": 0.02714073657989502, "chaser_new_fields": 94.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.78499984741211, "chaser_subspace_norm": 34.732913970947266, "collisions_per_episode": 1.350000023841858, "entropy": 0.0043078402522951365, "explorer_grad_norm": 0.7792868465185165, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 91.30493927001953, "final_distance": 3.708338975906372, "max_param_abs": 1.8683478832244873, "policy_loss": -0.000430340119405237, "value_loss": 0.09126785956323147}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:43.642889+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028623.6429017, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.9039363565347027, "elapsed_seconds": 6854.17088352493, "env_steps_per_second": 4085.103869718564, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.535881804993019, "eta_seconds": 12729.17449797487, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00018198393097622334, "chaser_grad_norm": 0.036112076602876186, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 36.714996337890625, "collisions_per_episode": 1.2000000476837158, "entropy": 0.003180745872668922, "explorer_grad_norm": 1.101660594344139, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 92.3774185180664, "final_distance": 3.520420789718628, "max_param_abs": 1.898836612701416, "policy_loss": -0.00017902375321154906, "value_loss": 0.09390992298722267}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:00.009138+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029120.009152, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.040748056849148, "elapsed_seconds": 7346.693004656932, "env_steps_per_second": 4083.4699341572536, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.4012467614152464, "eta_seconds": 12244.488341094888, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00042230356906974853, "chaser_grad_norm": 0.07996394485235214, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 37.123470306396484, "collisions_per_episode": 1.3250000476837158, "entropy": 0.0034858802682720125, "explorer_grad_norm": 0.5679749622941017, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 100.16153717041016, "final_distance": 3.3677127361297607, "max_param_abs": 1.9363983869552612, "policy_loss": -0.0003527298206460472, "value_loss": 0.09695837274193764}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:12.531257+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029612.5312696, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.176398255478028, "elapsed_seconds": 7835.033719720901, "env_steps_per_second": 4084.219818921201, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2645973832170423, "eta_seconds": 11752.550579581351, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0019244914414586756, "chaser_grad_norm": 0.12822234071791172, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.870000839233398, "chaser_subspace_norm": 41.26630783081055, "collisions_per_episode": 0.875, "entropy": 0.0027616055449470878, "explorer_grad_norm": 1.0141589939594269, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 100.3052749633789, "final_distance": 3.0593831539154053, "max_param_abs": 1.9546184539794922, "policy_loss": -0.0003915233662130646, "value_loss": 0.10390745662152767}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:20.871973+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030100.8719869, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.310681776522219, "elapsed_seconds": 8318.454395479988, "env_steps_per_second": 4087.297757919384, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1262165211771196, "eta_seconds": 11254.379476237631, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -1.4335217940003586e-05, "chaser_grad_norm": 0.04571158718317747, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.805000305175781, "chaser_subspace_norm": 41.69701385498047, "collisions_per_episode": 1.375, "entropy": 0.0025362682645209134, "explorer_grad_norm": 0.7047380357980728, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 105.92369079589844, "final_distance": 3.0623576641082764, "max_param_abs": 1.9880163669586182, "policy_loss": -8.808446089059174e-05, "value_loss": 0.1007557287812233}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:24.292648+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030584.2926602, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.4455032916399717, "elapsed_seconds": 8803.811849903897, "env_steps_per_second": 4089.1378204990806, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.988948467559965, "eta_seconds": 10760.214483215874, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00029666933588012867, "chaser_grad_norm": 0.040595004335045815, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.690000534057617, "chaser_subspace_norm": 42.999568939208984, "collisions_per_episode": 1.3000000715255737, "entropy": 0.0029416828765533864, "explorer_grad_norm": 0.41351157054305077, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 105.0737075805664, "final_distance": 2.9586522579193115, "max_param_abs": 2.0143489837646484, "policy_loss": -0.00024724579969159777, "value_loss": 0.11778833530843258}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:29.650106+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031069.6501222, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.5814078495824813, "elapsed_seconds": 9293.068258496933, "env_steps_per_second": 4089.069287235188, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.853134991643795, "eta_seconds": 10271.285969917662, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0017763836128132163, "chaser_grad_norm": 0.06757924798876047, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.824999809265137, "chaser_subspace_norm": 42.66282272338867, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0017758093017619103, "explorer_grad_norm": 1.1538244783878326, "explorer_new_fields": 97.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 109.50704956054688, "final_distance": 3.04056715965271, "max_param_abs": 2.048684597015381, "policy_loss": -0.0013928900831245983, "value_loss": 0.11639217287302017}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:38.906511+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031558.9065232, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.718024756178331, "elapsed_seconds": 9784.889122241992, "env_steps_per_second": 4087.935949021248, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.718024756178331, "eta_seconds": 9784.889122241992, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -1.4062203634868098e-05, "chaser_grad_norm": 0.05896423850208521, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 46.2486572265625, "collisions_per_episode": 1.0750000476837158, "entropy": 0.003515815536957234, "explorer_grad_norm": 0.31926171854138374, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 108.01021575927734, "final_distance": 2.9094836711883545, "max_param_abs": 2.0774147510528564, "policy_loss": -0.000163217312160846, "value_loss": 0.08966618403792381}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:50.727376+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032050.7273905, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.853866963133882, "elapsed_seconds": 10273.921067281975, "env_steps_per_second": 4088.0205059927857, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5820701095020837, "eta_seconds": 9295.452394207501, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0030544491095406134, "chaser_grad_norm": 0.15306414104998112, "chaser_new_fields": 94.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.885001182556152, "chaser_subspace_norm": 48.492122650146484, "collisions_per_episode": 1.024999976158142, "entropy": 0.00357656559208408, "explorer_grad_norm": 0.9481620192527771, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 113.7264633178711, "final_distance": 3.156186103820801, "max_param_abs": 2.099663019180298, "policy_loss": -0.0007126822820868739, "value_loss": 0.11961465328931808}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:59.759321+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032539.7593343, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.9875429796324977, "elapsed_seconds": 10755.154726676992, "env_steps_per_second": 4091.0615531046506, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4443533469720435, "eta_seconds": 8799.672049099358, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00020237287433322315, "chaser_grad_norm": 0.03940554801374674, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 48.79186248779297, "collisions_per_episode": 1.3000000715255737, "entropy": 0.0031767572509124875, "explorer_grad_norm": 0.48446035385131836, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 116.57722473144531, "final_distance": 3.3267982006073, "max_param_abs": 2.1329665184020996, "policy_loss": -0.000321724902139664, "value_loss": 0.10522451065480709}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:00.992980+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033020.992996, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.121035432090252, "elapsed_seconds": 11235.727555524907, "env_steps_per_second": 4094.0828951820367, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.306852275892795, "eta_seconds": 8304.668193214062, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004090758606384304, "chaser_grad_norm": 0.19424590468406677, "chaser_new_fields": 94.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 50.733299255371094, "collisions_per_episode": 1.25, "entropy": 0.002097469550790265, "explorer_grad_norm": 0.6892369091510773, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 116.6259765625, "final_distance": 2.8124120235443115, "max_param_abs": 2.1573104858398438, "policy_loss": -0.0005452695031418209, "value_loss": 0.10252008028328419}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:01.565808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033501.5658207, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.254765800418035, "elapsed_seconds": 11717.156881504925, "env_steps_per_second": 4096.556910982913, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1698438669453566, "eta_seconds": 7811.437921003284, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.001365038953287545, "chaser_grad_norm": 0.10101035609841347, "chaser_new_fields": 94.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 50.75255584716797, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0037602168740704656, "explorer_grad_norm": 0.43663299828767776, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 119.86353302001953, "final_distance": 3.1305291652679443, "max_param_abs": 2.1793112754821777, "policy_loss": -6.250381696659701e-05, "value_loss": 0.10334442555904388}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:02.995136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033982.995153, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.3898560839169334, "elapsed_seconds": 12203.48190210096, "env_steps_per_second": 4097.191309915571, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0339136503501605, "eta_seconds": 7322.089141260577, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.000989114510704936, "chaser_grad_norm": 0.08919528219848871, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 51.018123626708984, "collisions_per_episode": 1.274999976158142, "entropy": 0.0036947763292118907, "explorer_grad_norm": 0.49894385039806366, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 125.3757095336914, "final_distance": 2.5259768962860107, "max_param_abs": 2.2026240825653076, "policy_loss": -0.0006352567888097127, "value_loss": 0.11523559875786304}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:21:09.320156+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034469.3201694, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5251186574160887, "elapsed_seconds": 12690.427166697918, "env_steps_per_second": 4097.576804700305, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8981408155317399, "eta_seconds": 6833.306935914264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 9.058276429174672e-05, "chaser_grad_norm": 0.048134016804397106, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 52.3425178527832, "collisions_per_episode": 1.5, "entropy": 0.0022945620585232973, "explorer_grad_norm": 1.0205857157707214, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 129.28347778320312, "final_distance": 2.797969102859497, "max_param_abs": 2.2236173152923584, "policy_loss": -8.505058767127593e-05, "value_loss": 0.11947360821068287}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:29:16.265420+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034956.265435, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.6616521055691473, "elapsed_seconds": 13181.94758004893, "env_steps_per_second": 4096.511511070624, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7630176804592192, "eta_seconds": 6346.863649653189, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00023019519146316725, "chaser_grad_norm": 0.04138797149062157, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 56.570579528808594, "collisions_per_episode": 1.2000000476837158, "entropy": 0.002282455679960549, "explorer_grad_norm": 0.938120424747467, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.3125, "explorer_subspace_norm": 131.00253295898438, "final_distance": 2.6901204586029053, "max_param_abs": 2.2481918334960938, "policy_loss": -0.00024209464679003645, "value_loss": 0.11030559986829758}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:37:27.785834+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035447.7858477, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.797376491899712, "elapsed_seconds": 13670.555370838963, "env_steps_per_second": 4096.395390011377, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6274470679570197, "eta_seconds": 5858.809444645271, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0015203803531047555, "chaser_grad_norm": 0.07362508960068226, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.690000534057617, "chaser_subspace_norm": 58.064247131347656, "collisions_per_episode": 1.475000023841858, "entropy": 0.004419806529767811, "explorer_grad_norm": 1.720163807272911, "explorer_new_fields": 97.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 126.17747497558594, "final_distance": 2.929892063140869, "max_param_abs": 2.2679409980773926, "policy_loss": -0.0007249031414975349, "value_loss": 0.13401664420962334}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:45:36.393624+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035936.3936381, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.93200999457166, "elapsed_seconds": 14155.235980457976, "env_steps_per_second": 4097.423743417062, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4914520669064917, "eta_seconds": 5369.22744086337, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0016421328080014064, "chaser_grad_norm": 0.1113754017278552, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 60.42548370361328, "collisions_per_episode": 1.024999976158142, "entropy": 0.003006738726980984, "explorer_grad_norm": 0.8497034683823586, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 130.97738647460938, "final_distance": 2.5862414836883545, "max_param_abs": 2.383998394012451, "policy_loss": -0.0005429268329173453, "value_loss": 0.11908523179590702}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:53:41.074235+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036421.0742488, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.069114904249695, "elapsed_seconds": 14648.813655298902, "env_steps_per_second": 4095.8948220558636, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3563716347498982, "eta_seconds": 4882.937885099634, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004713547696755249, "chaser_grad_norm": 0.031762602273374796, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 60.8377799987793, "collisions_per_episode": 1.149999976158142, "entropy": 0.0015576633741147816, "explorer_grad_norm": 0.5975871719419956, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 135.2548065185547, "final_distance": 2.847369432449341, "max_param_abs": 2.522535562515259, "policy_loss": -0.00016278053342588805, "value_loss": 0.07293580658733845}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:01:54.651910+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036914.6519248, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.205038526879992, "elapsed_seconds": 15138.138696767972, "env_steps_per_second": 4095.6157980793996, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2208176368361268, "eta_seconds": 4394.943492610057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004081863581145151, "chaser_grad_norm": 0.05001601576805115, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.84000015258789, "chaser_subspace_norm": 59.43809509277344, "collisions_per_episode": 0.7750000357627869, "entropy": 0.0034388050553388894, "explorer_grad_norm": 0.8913935273885727, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 139.73411560058594, "final_distance": 3.0137009620666504, "max_param_abs": 2.671187400817871, "policy_loss": -0.00016906086285572997, "value_loss": 0.10254292003810406}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:10:03.976952+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037403.976966, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.3405951427749825, "elapsed_seconds": 15626.142513989937, "env_steps_per_second": 4095.700518710962, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0851487856937456, "eta_seconds": 3906.535628497484, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0013712222712034539, "chaser_grad_norm": 0.06847671559080482, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.74000072479248, "chaser_subspace_norm": 59.01997756958008, "collisions_per_episode": 1.0500000715255737, "entropy": 0.003109973040409386, "explorer_grad_norm": 0.5139616504311562, "explorer_new_fields": 97.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 139.3917236328125, "final_distance": 2.5476038455963135, "max_param_abs": 2.8179855346679688, "policy_loss": -0.00032580759574152296, "value_loss": 0.09241721779108047}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:18:11.980771+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037891.9807906, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.476565459541388, "elapsed_seconds": 16115.635654348996, "env_steps_per_second": 4095.4015972797893, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9495744914178701, "eta_seconds": 3418.4681691043324, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0002272066812404605, "chaser_grad_norm": 0.03070598654448986, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 62.595947265625, "collisions_per_episode": 0.9750000238418579, "entropy": 0.002077111799735576, "explorer_grad_norm": 0.3588371053338051, "explorer_new_fields": 97.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 132.83013916015625, "final_distance": 3.009052038192749, "max_param_abs": 2.94905161857605, "policy_loss": -7.086706634318318e-05, "value_loss": 0.0682746060192585}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:26:21.473906+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038381.4739184, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.613364270519976, "elapsed_seconds": 16608.111373871914, "env_steps_per_second": 4094.3848743077697, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8141231065623488, "eta_seconds": 2930.8431836244554, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0020441413901413075, "chaser_grad_norm": 0.045831756200641394, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 65.60694885253906, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0020277429139241576, "explorer_grad_norm": 0.5332756787538528, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 145.29444885253906, "final_distance": 3.0503880977630615, "max_param_abs": 3.096259832382202, "policy_loss": -6.159306025033118e-05, "value_loss": 0.07014592923223972}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:33.949628+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038873.9496448, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.749466954409694, "elapsed_seconds": 17098.0810358749, "env_steps_per_second": 4094.0266836452115, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.678495279201385, "eta_seconds": 2442.583005124986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0002671154577380541, "chaser_grad_norm": 0.0547892265021801, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 66.88868713378906, "collisions_per_episode": 0.949999988079071, "entropy": 0.0021956649725325406, "explorer_grad_norm": 0.6824228912591934, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 147.3888702392578, "final_distance": 2.673384666442871, "max_param_abs": 3.2412221431732178, "policy_loss": -0.000199659829901222, "value_loss": 0.08626480400562286}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:43.919290+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039363.9193044, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.8856042489344444, "elapsed_seconds": 17588.175296164, "env_steps_per_second": 4093.659449465646, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428449165482716, "eta_seconds": 1954.2416995737779, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00024259486997357804, "chaser_grad_norm": 0.06348314369097352, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 68.30133819580078, "collisions_per_episode": 1.225000023841858, "entropy": 0.0032776488224044442, "explorer_grad_norm": 0.3755941763520241, "explorer_new_fields": 97.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 144.0692901611328, "final_distance": 2.865546464920044, "max_param_abs": 3.3910720348358154, "policy_loss": 1.6978267027401017e-05, "value_loss": 0.09523412771522999}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:50:54.013551+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039854.0135665, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.021417094115556, "elapsed_seconds": 18077.101538816, "env_steps_per_second": 4093.576608014494, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40714192654990994, "eta_seconds": 1465.7109355796758, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 7.123219962856808e-05, "chaser_grad_norm": 0.036403757985681295, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 70.68766784667969, "collisions_per_episode": 1.225000023841858, "entropy": 0.0030727192061021924, "explorer_grad_norm": 0.42910633981227875, "explorer_new_fields": 97.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 145.63185119628906, "final_distance": 2.8040151596069336, "max_param_abs": 3.5327699184417725, "policy_loss": -0.0001866714523055535, "value_loss": 0.10456709191203117}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:59:02.939791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040342.9398034, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.155799050785556, "elapsed_seconds": 18560.876582828, "env_steps_per_second": 4094.6341979512463, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27135784477818714, "eta_seconds": 976.8882412014738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0011629241046633432, "chaser_grad_norm": 0.09430215880274773, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 69.79297637939453, "collisions_per_episode": 1.225000023841858, "entropy": 0.0034293356584385037, "explorer_grad_norm": 0.4956182390451431, "explorer_new_fields": 97.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 145.0971221923828, "final_distance": 3.0578739643096924, "max_param_abs": 3.6705596446990967, "policy_loss": -0.0006123671865765346, "value_loss": 0.08534514717757702}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:06.714836+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040826.7148514, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.29110973319637, "elapsed_seconds": 19047.99503950693, "env_steps_per_second": 4094.919168039592, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13566948033836845, "eta_seconds": 488.41012921812643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0001795099746915163, "chaser_grad_norm": 0.05988466367125511, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 70.86493682861328, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0021517517743632197, "explorer_grad_norm": 0.41555262356996536, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 155.8712921142578, "final_distance": 3.2743000984191895, "max_param_abs": 3.82509183883667, "policy_loss": -6.89809348913828e-05, "value_loss": 0.062252127565443516}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:15:13.833292+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041313.8333035, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.423251158835806, "elapsed_seconds": 19523.704171808902, "env_steps_per_second": 4097.583086488033, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:09.542424+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041789.5424345, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.423259738547204, "elapsed_seconds": 19523.735058769933, "env_steps_per_second": 4097.5766040250855, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:09.573312+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041789.573325, "update": 20000}

non_social/seed_0005/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.423259738547204,
+  "elapsed_seconds": 19523.735058769933,
+  "env_steps_per_second": 4097.5766040250855,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 4.4411225102368446e-05,
+    "chaser_grad_norm": 0.09301618672907352,
+    "chaser_new_fields": 94.4000015258789,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.815000534057617,
+    "chaser_subspace_norm": 69.54359436035156,
+    "collisions_per_episode": 0.925000011920929,
+    "entropy": 0.002813033468555659,
+    "explorer_grad_norm": 0.6812433376908302,
+    "explorer_new_fields": 97.95000457763672,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 96.0625,
+    "explorer_subspace_norm": 156.21910095214844,
+    "final_distance": 3.1009461879730225,
+    "max_param_abs": 3.969390392303467,
+    "policy_loss": -0.0003465476266564238,
+    "value_loss": 0.07470510713756084
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-03T01:23:09.573312+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783041789.573325,
+  "update": 20000
+}

non_social/seed_0006/attempt_01/checkpoints/latest.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3eb7fb2c91b2c029361c2a8a5a023ab29aba5b6cae2aca1cd33fac54dfe1d6a
+size 29760368

non_social/seed_0006/attempt_01/config.json

@@ -0,0 +1,44 @@
+{
+  "architecture": "transformer",
+  "attempt_id": "attempt_01",
+  "batch_size": 40,
+  "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+  "checkpoint_every": 1000,
+  "checkpoint_every_seconds": 1800.0,
+  "clip_epsilon": 0.2,
+  "cost_per_hour": null,
+  "cuda_tf32": true,
+  "device": "cuda",
+  "entropy_coef": 0.01,
+  "env": {
+    "grid_size": 10,
+    "max_steps": 100,
+    "partner_visibility": "none",
+    "task": "non_social",
+    "vision_radius": 3
+  },
+  "experiment_id": "mouse-run-run-2-transformer",
+  "finite_guard": true,
+  "fused_agent_rollout": false,
+  "gae_lambda": 0.95,
+  "gamma": 0.99,
+  "grad_clip": 1.0,
+  "hidden_size": 256,
+  "learning_rate": 0.0003,
+  "log_every": 500,
+  "metrics_path": null,
+  "ppo_epochs": 4,
+  "recurrent_l2_coef": 0.3,
+  "resume_from": null,
+  "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01",
+  "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01",
+  "seed": 6,
+  "status_every_seconds": 1800.0,
+  "status_path": null,
+  "subspace_metric_period": 10,
+  "tensorboard_dir": null,
+  "triton_env_step": true,
+  "updates": 20000,
+  "value_clip": 10.0,
+  "value_coef": 0.5
+}

non_social/seed_0006/attempt_01/metrics.jsonl

@@ -0,0 +1,42 @@
+{"device": "cuda", "elapsed_hours": 0.0006340497011034232, "elapsed_seconds": 2.2825789239723235, "env_steps_per_second": 1752.4038086879752, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 12.680359972367361, "eta_seconds": 45649.2959005225, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.009081356627168358, "chaser_grad_norm": 9.28392294794321, "chaser_new_fields": 34.45000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.135000228881836, "chaser_subspace_norm": 0.032425787299871445, "collisions_per_episode": 1.350000023841858, "entropy": 2.7488085627555847, "explorer_grad_norm": 7.50808984041214, "explorer_new_fields": 32.82500076293945, "explorer_partner_vision": 0.0, "explorer_return": -0.949999988079071, "explorer_subspace_norm": 0.0343438945710659, "final_distance": 5.18532657623291, "max_param_abs": 1.0012028217315674, "policy_loss": -0.008591297162709566, "value_loss": 7.798083543777466}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:10.457463+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022290.4574733, "update": 1}
+{"device": "cuda", "elapsed_hours": 0.1324592853458469, "elapsed_seconds": 476.8534272450488, "env_steps_per_second": 4194.160900876205, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.165912128488029, "eta_seconds": 18597.283662556903, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0017517860296626386, "chaser_grad_norm": 0.31775665283203125, "chaser_new_fields": 91.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 11.57347297668457, "collisions_per_episode": 0.75, "entropy": 0.0558307059109211, "explorer_grad_norm": 10.75618839263916, "explorer_new_fields": 93.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.4375, "explorer_subspace_norm": 15.667323112487793, "final_distance": 2.812631368637085, "max_param_abs": 1.32639741897583, "policy_loss": -0.0034992986684585503, "value_loss": 0.35839923471212387}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:05.028314+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022765.028328, "update": 500}
+{"device": "cuda", "elapsed_hours": 0.26846497917001316, "elapsed_seconds": 966.4739250120474, "env_steps_per_second": 4138.756252477416, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.100834604230251, "eta_seconds": 18363.0045752289, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0014362353670867023, "chaser_grad_norm": 0.4267645925283432, "chaser_new_fields": 92.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 13.562819480895996, "collisions_per_episode": 0.5, "entropy": 0.018285356927663088, "explorer_grad_norm": 4.223830282688141, "explorer_new_fields": 96.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 28.409225463867188, "final_distance": 2.7266013622283936, "max_param_abs": 1.5229872465133667, "policy_loss": -0.002394474568310656, "value_loss": 0.1739034280180931}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:14:14.648810+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023254.648824, "update": 1000}
+{"device": "cuda", "elapsed_hours": 0.4027435065277839, "elapsed_seconds": 1449.876623500022, "env_steps_per_second": 4138.28314957994, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.967169913842668, "eta_seconds": 17881.811689833605, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004477103975858654, "chaser_grad_norm": 0.32416776195168495, "chaser_new_fields": 90.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.040000915527344, "chaser_subspace_norm": 16.41919708251953, "collisions_per_episode": 0.949999988079071, "entropy": 0.01877806894481182, "explorer_grad_norm": 3.8259685933589935, "explorer_new_fields": 96.625, "explorer_partner_vision": 0.0, "explorer_return": 94.4124984741211, "explorer_subspace_norm": 39.73762893676758, "final_distance": 2.208803176879883, "max_param_abs": 1.6338715553283691, "policy_loss": -0.0017088156484490469, "value_loss": 0.17939159274101257}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:18.051506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023738.0515182, "update": 1500}
+{"device": "cuda", "elapsed_hours": 0.5402188905961035, "elapsed_seconds": 1944.7880061459728, "env_steps_per_second": 4113.558894192158, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.861970015364933, "eta_seconds": 17503.09205531376, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0007497676800674591, "chaser_grad_norm": 0.1537339510396123, "chaser_new_fields": 92.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.440001487731934, "chaser_subspace_norm": 20.747173309326172, "collisions_per_episode": 0.699999988079071, "entropy": 0.008710291935130954, "explorer_grad_norm": 2.2933662235736847, "explorer_new_fields": 96.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 46.05669021606445, "final_distance": 2.3678863048553467, "max_param_abs": 1.6998072862625122, "policy_loss": -0.0007282829068131491, "value_loss": 0.10590462386608124}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:32.962892+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024232.9629066, "update": 2000}
+{"device": "cuda", "elapsed_hours": 0.6760303580455689, "elapsed_seconds": 2433.709288964048, "env_steps_per_second": 4108.954198164185, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.732212506318982, "eta_seconds": 17035.965022748336, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002755190706777988, "chaser_grad_norm": 0.10564791224896908, "chaser_new_fields": 93.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.520000457763672, "chaser_subspace_norm": 23.54489517211914, "collisions_per_episode": 1.0500000715255737, "entropy": 0.009279369376599789, "explorer_grad_norm": 2.4777910113334656, "explorer_new_fields": 96.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.5250015258789, "explorer_subspace_norm": 52.84782028198242, "final_distance": 2.5959105491638184, "max_param_abs": 1.7529109716415405, "policy_loss": -0.0013189362334102839, "value_loss": 0.14943654462695122}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:41.884172+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024721.8841834, "update": 2500}
+{"device": "cuda", "elapsed_hours": 0.8094591433922242, "elapsed_seconds": 2914.0529162120074, "env_steps_per_second": 4117.976009714629, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.58693514588927, "eta_seconds": 16512.966525201373, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0026490730582038324, "chaser_grad_norm": 0.1542376298457384, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 28.20026397705078, "collisions_per_episode": 0.6500000357627869, "entropy": 0.005760655738413334, "explorer_grad_norm": 2.126728057861328, "explorer_new_fields": 96.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 61.04405212402344, "final_distance": 3.0504865646362305, "max_param_abs": 1.7990180253982544, "policy_loss": -0.0018209304869998277, "value_loss": 0.1942107118666172}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:42.227800+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025202.227815, "update": 3000}
+{"device": "cuda", "elapsed_hours": 0.9441227549747175, "elapsed_seconds": 3398.8419179089833, "env_steps_per_second": 4119.050058266023, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.450864416309383, "eta_seconds": 16023.111898713778, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.002258445791657515, "chaser_grad_norm": 0.18437314964830875, "chaser_new_fields": 93.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 29.267290115356445, "collisions_per_episode": 0.949999988079071, "entropy": 0.006399786798283458, "explorer_grad_norm": 0.8159359097480774, "explorer_new_fields": 96.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.0374984741211, "explorer_subspace_norm": 63.45840835571289, "final_distance": 2.536119222640991, "max_param_abs": 1.8377091884613037, "policy_loss": -0.001426826559240979, "value_loss": 0.11861684918403625}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:47.016801+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025687.016813, "update": 3500}
+{"device": "cuda", "elapsed_hours": 1.0790977155599912, "elapsed_seconds": 3884.751776015968, "env_steps_per_second": 4118.66727207186, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.316390862239965, "eta_seconds": 15539.007104063872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0009261028763427048, "chaser_grad_norm": 0.07158619537949562, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 31.17525863647461, "collisions_per_episode": 0.75, "entropy": 0.006486838683485985, "explorer_grad_norm": 1.674952894449234, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 68.22615051269531, "final_distance": 2.0703301429748535, "max_param_abs": 1.878548264503479, "policy_loss": -0.0013943439025752014, "value_loss": 0.15084588155150414}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:02:52.926661+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026172.9266772, "update": 4000}
+{"device": "cuda", "elapsed_hours": 1.2143827589272083, "elapsed_seconds": 4371.777932137949, "env_steps_per_second": 4117.318006406008, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.182873947415939, "eta_seconds": 15058.34621069738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 6.878799733930907e-05, "chaser_grad_norm": 0.03787218267098069, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 34.393680572509766, "collisions_per_episode": 1.024999976158142, "entropy": 0.004154657945036888, "explorer_grad_norm": 1.4655900299549103, "explorer_new_fields": 96.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 72.24091339111328, "final_distance": 2.4027812480926514, "max_param_abs": 1.914402723312378, "policy_loss": -0.00025126971033806456, "value_loss": 0.09570373967289925}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:10:59.952815+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026659.9528258, "update": 4500}
+{"device": "cuda", "elapsed_hours": 1.351526881251096, "elapsed_seconds": 4865.496772503946, "env_steps_per_second": 4110.577179502955, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.0545806437532885, "eta_seconds": 14596.490317511838, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 1.0621170141389413e-05, "chaser_grad_norm": 0.0821846779435873, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 33.195072174072266, "collisions_per_episode": 0.8500000238418579, "entropy": 0.004661013721488416, "explorer_grad_norm": 1.5245185494422913, "explorer_new_fields": 97.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 74.63639831542969, "final_distance": 2.5133373737335205, "max_param_abs": 1.9484796524047852, "policy_loss": -0.0004091162831763029, "value_loss": 0.06395275145769119}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:13.671657+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027153.6716738, "update": 5000}
+{"device": "cuda", "elapsed_hours": 1.4883398381705693, "elapsed_seconds": 5358.02341741405, "env_steps_per_second": 4105.991759666084, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9238050279042285, "eta_seconds": 14125.698100455222, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00039155774840083507, "chaser_grad_norm": 0.04648858401924372, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.610000610351562, "chaser_subspace_norm": 36.63489532470703, "collisions_per_episode": 0.925000011920929, "entropy": 0.0043410195503383875, "explorer_grad_norm": 1.281730055809021, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 80.35256958007812, "final_distance": 2.188136339187622, "max_param_abs": 1.9779452085494995, "policy_loss": -0.0006414571033133143, "value_loss": 0.07625494711101055}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:26.198302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027646.1983166, "update": 5500}
+{"device": "cuda", "elapsed_hours": 1.6236991062505532, "elapsed_seconds": 5845.316782501992, "env_steps_per_second": 4105.851041614069, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7886312479179574, "eta_seconds": 13639.072492504647, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0023648358524350765, "chaser_grad_norm": 0.035658171866089106, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 40.32418441772461, "collisions_per_episode": 0.574999988079071, "entropy": 0.004197318048682064, "explorer_grad_norm": 1.142631858587265, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 87.25827026367188, "final_distance": 2.5605618953704834, "max_param_abs": 2.0082149505615234, "policy_loss": -0.0031194363399400515, "value_loss": 0.11602420918643475}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:33.491668+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028133.4916828, "update": 6000}
+{"device": "cuda", "elapsed_hours": 1.7596862807822051, "elapsed_seconds": 6334.870610815939, "env_steps_per_second": 4104.266937292847, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6547330447015036, "eta_seconds": 13157.038960925413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003027150061973849, "chaser_grad_norm": 0.04697199258953333, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 39.59521484375, "collisions_per_episode": 1.2000000476837158, "entropy": 0.006377081153914332, "explorer_grad_norm": 0.44872088730335236, "explorer_new_fields": 96.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.75, "explorer_subspace_norm": 88.96295928955078, "final_distance": 2.341438055038452, "max_param_abs": 2.0350699424743652, "policy_loss": -0.00037196256092286717, "value_loss": 0.1247739065438509}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:43.045494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028623.0455053, "update": 6500}
+{"device": "cuda", "elapsed_hours": 1.8960810574647298, "elapsed_seconds": 6825.891806873027, "env_steps_per_second": 4102.028100094796, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5212933924344982, "eta_seconds": 12676.656212764194, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0015827478717496124, "chaser_grad_norm": 0.1372872330248356, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.675000190734863, "chaser_subspace_norm": 41.84708786010742, "collisions_per_episode": 0.6500000357627869, "entropy": 0.003968516713939607, "explorer_grad_norm": 0.7732554823160172, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.61250305175781, "explorer_subspace_norm": 88.53568267822266, "final_distance": 2.46395206451416, "max_param_abs": 2.065558433532715, "policy_loss": -0.001052412109914691, "value_loss": 0.0816511232405901}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:51:54.066690+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029114.0667057, "update": 7000}
+{"device": "cuda", "elapsed_hours": 2.0298366709441567, "elapsed_seconds": 7307.412015398964, "env_steps_per_second": 4105.420624535851, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3830611182402617, "eta_seconds": 12179.020025664942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0019360221052711346, "chaser_grad_norm": 0.05764288455247879, "chaser_new_fields": 93.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 43.68484878540039, "collisions_per_episode": 0.699999988079071, "entropy": 0.0032946523278951645, "explorer_grad_norm": 0.9634853154420853, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 96.0250015258789, "explorer_subspace_norm": 93.03872680664062, "final_distance": 2.0583181381225586, "max_param_abs": 2.0885467529296875, "policy_loss": -0.00018088604579613587, "value_loss": 0.06554706953465939}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:59:55.586898+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029595.5869105, "update": 7500}
+{"device": "cuda", "elapsed_hours": 2.164390133535554, "elapsed_seconds": 7791.804480727995, "env_steps_per_second": 4106.879231781008, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2465852003033313, "eta_seconds": 11687.706721091992, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005512295202090844, "chaser_grad_norm": 0.0987094221636653, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 48.18280792236328, "collisions_per_episode": 0.949999988079071, "entropy": 0.0032457555062137544, "explorer_grad_norm": 0.5241964012384415, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 93.68035125732422, "final_distance": 2.1907718181610107, "max_param_abs": 2.120203733444214, "policy_loss": -0.000570195695850817, "value_loss": 0.08561011590063572}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:07:59.979365+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030079.9793801, "update": 8000}
+{"device": "cuda", "elapsed_hours": 2.297568872478894, "elapsed_seconds": 8271.247940924019, "env_steps_per_second": 4110.625173231321, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.108475533353798, "eta_seconds": 11190.511920073672, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 9.063592311420177e-05, "chaser_grad_norm": 0.03654113505035639, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.690001487731934, "chaser_subspace_norm": 48.02528381347656, "collisions_per_episode": 1.024999976158142, "entropy": 0.004057165002450347, "explorer_grad_norm": 0.4029681459069252, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 96.40164947509766, "final_distance": 2.4163246154785156, "max_param_abs": 2.141502857208252, "policy_loss": -0.00027873397825173285, "value_loss": 0.11679401993751526}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:15:59.422823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030559.422835, "update": 8500}
+{"device": "cuda", "elapsed_hours": 2.432162019604455, "elapsed_seconds": 8755.783270576037, "env_steps_per_second": 4111.568192988357, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9726424684054447, "eta_seconds": 10701.512886259601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": -2.4071496259894798e-05, "chaser_grad_norm": 0.025664323708042502, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 51.08539962768555, "collisions_per_episode": 0.8500000238418579, "entropy": 0.002273507008794695, "explorer_grad_norm": 0.4124542959034443, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 96.16495513916016, "final_distance": 2.28855562210083, "max_param_abs": 2.1810531616210938, "policy_loss": -5.639028652515421e-05, "value_loss": 0.1025311816483736}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:03.958154+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031043.9581697, "update": 9000}
+{"device": "cuda", "elapsed_hours": 2.56710545852056, "elapsed_seconds": 9241.579650674015, "env_steps_per_second": 4111.851159258098, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8373270857332504, "eta_seconds": 10214.377508639702, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00022325854453769667, "chaser_grad_norm": 0.019994712667539716, "chaser_new_fields": 94.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 53.11989212036133, "collisions_per_episode": 1.024999976158142, "entropy": 0.001609528815606609, "explorer_grad_norm": 0.5562155470252037, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.1624984741211, "explorer_subspace_norm": 97.70769500732422, "final_distance": 2.372077703475952, "max_param_abs": 2.2171342372894287, "policy_loss": -0.00012136198038970036, "value_loss": 0.11478429846465588}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:09.754534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031529.754543, "update": 9500}
+{"device": "cuda", "elapsed_hours": 2.7005695298522348, "elapsed_seconds": 9722.050307468046, "env_steps_per_second": 4114.358467089373, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7005695298522348, "eta_seconds": 9722.050307468046, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.002776364052704955, "chaser_grad_norm": 0.06105522019788623, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.65999984741211, "chaser_subspace_norm": 53.78749084472656, "collisions_per_episode": 1.0, "entropy": 0.0024248933186754584, "explorer_grad_norm": 0.4683804363012314, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6875, "explorer_subspace_norm": 97.96976470947266, "final_distance": 2.3905584812164307, "max_param_abs": 2.239759922027588, "policy_loss": -0.0006870093730952931, "value_loss": 0.07492147944867611}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:10.225191+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032010.2252047, "update": 10000}
+{"device": "cuda", "elapsed_hours": 2.835848784226659, "elapsed_seconds": 10209.055623215972, "env_steps_per_second": 4113.994628894921, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5657679476336437, "eta_seconds": 9236.764611481118, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 5.5746936133201086e-05, "chaser_grad_norm": 0.0356141347438097, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 55.35414123535156, "collisions_per_episode": 1.024999976158142, "entropy": 0.003112861479166895, "explorer_grad_norm": 0.4594806991517544, "explorer_new_fields": 97.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 100.5293960571289, "final_distance": 2.288546323776245, "max_param_abs": 2.2687058448791504, "policy_loss": -3.996277139961535e-05, "value_loss": 0.09902623109519482}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:17.230506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032497.2305174, "update": 10500}
+{"device": "cuda", "elapsed_hours": 2.971485036740002, "elapsed_seconds": 10697.346132264007, "env_steps_per_second": 4113.169701716266, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.431215030060002, "eta_seconds": 8752.374108216007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00013370348195351767, "chaser_grad_norm": 0.020539223216474056, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 54.3487434387207, "collisions_per_episode": 0.8500000238418579, "entropy": 0.003950043930672109, "explorer_grad_norm": 0.31602199375629425, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 107.02349090576172, "final_distance": 2.685612916946411, "max_param_abs": 2.2876455783843994, "policy_loss": -0.00022414303993123497, "value_loss": 0.07805249281227589}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:56:25.521017+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032985.521044, "update": 11000}
+{"device": "cuda", "elapsed_hours": 3.1068254964366657, "elapsed_seconds": 11184.571787171997, "env_steps_per_second": 4112.808328769378, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.296349279974927, "eta_seconds": 8266.857407909738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": -6.989048116023966e-06, "chaser_grad_norm": 0.03843334084376693, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 54.69615936279297, "collisions_per_episode": 0.949999988079071, "entropy": 0.002638183650560677, "explorer_grad_norm": 0.32337915152311325, "explorer_new_fields": 97.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 104.26318359375, "final_distance": 2.929190158843994, "max_param_abs": 2.3113341331481934, "policy_loss": -0.00012264920040583327, "value_loss": 0.09451063349843025}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:04:32.746670+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033472.746682, "update": 11500}
+{"device": "cuda", "elapsed_hours": 3.241158140361949, "elapsed_seconds": 11668.169305303018, "env_steps_per_second": 4113.755872413051, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.160772093574633, "eta_seconds": 7778.779536868678, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006010586898232972, "chaser_grad_norm": 0.09262969810515642, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 58.6978759765625, "collisions_per_episode": 0.7750000357627869, "entropy": 0.005693767685443163, "explorer_grad_norm": 0.5575678050518036, "explorer_new_fields": 96.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.5999984741211, "explorer_subspace_norm": 107.43771362304688, "final_distance": 2.5675151348114014, "max_param_abs": 2.3327088356018066, "policy_loss": -0.0006506353926347686, "value_loss": 0.14505011588335037}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:12:36.344190+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033956.344204, "update": 12000}
+{"device": "cuda", "elapsed_hours": 3.376459176991387, "elapsed_seconds": 12155.253037168994, "env_steps_per_second": 4113.447893442225, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.025875506194832, "eta_seconds": 7293.151822301395, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0007247018451639331, "chaser_grad_norm": 0.03711062157526612, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.550000190734863, "chaser_subspace_norm": 60.891971588134766, "collisions_per_episode": 1.149999976158142, "entropy": 0.003253880888223648, "explorer_grad_norm": 0.3885218985378742, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 105.75569152832031, "final_distance": 3.1141555309295654, "max_param_abs": 2.343086004257202, "policy_loss": -0.00020295001997094886, "value_loss": 0.10075781308114529}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:20:43.427921+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034443.4279327, "update": 12500}
+{"device": "cuda", "elapsed_hours": 3.5124197246261044, "elapsed_seconds": 12644.711008653976, "env_steps_per_second": 4112.391336141369, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8913029286448255, "eta_seconds": 6808.690543121372, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.000679097149712593, "chaser_grad_norm": 0.03848004946485162, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 58.41977310180664, "collisions_per_episode": 1.024999976158142, "entropy": 0.003096371015999466, "explorer_grad_norm": 0.3082907684147358, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 111.32701110839844, "final_distance": 2.35443115234375, "max_param_abs": 2.37701678276062, "policy_loss": -0.0004093637610793621, "value_loss": 0.07662378810346127}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:28:52.885893+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034932.8859081, "update": 13000}
+{"device": "cuda", "elapsed_hours": 3.6460182049038767, "elapsed_seconds": 13125.665537653957, "env_steps_per_second": 4114.077099183178, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7554902468055704, "eta_seconds": 6319.7648885000535, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -4.818960044783438e-05, "chaser_grad_norm": 0.046644340734928846, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.780000686645508, "chaser_subspace_norm": 59.496761322021484, "collisions_per_episode": 0.7750000357627869, "entropy": 0.002635331475175917, "explorer_grad_norm": 0.36724893748760223, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 115.20111083984375, "final_distance": 2.6180009841918945, "max_param_abs": 2.387871026992798, "policy_loss": -0.0002843060591501967, "value_loss": 0.06970574520528316}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:36:53.840420+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035413.8404324, "update": 13500}
+{"device": "cuda", "elapsed_hours": 3.7837413430186118, "elapsed_seconds": 13621.468834867002, "env_steps_per_second": 4111.157223856525, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6216034327222624, "eta_seconds": 5837.772357800144, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -4.5233418649925916e-05, "chaser_grad_norm": 0.031938156113028526, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.734999656677246, "chaser_subspace_norm": 64.35379791259766, "collisions_per_episode": 0.9000000357627869, "entropy": 0.00153730483725667, "explorer_grad_norm": 0.3889255002140999, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 114.7887191772461, "final_distance": 2.327620267868042, "max_param_abs": 2.4029719829559326, "policy_loss": -2.1636963527527087e-05, "value_loss": 0.07703442126512527}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:45:09.643720+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035909.6437354, "update": 14000}
+{"device": "cuda", "elapsed_hours": 3.9194211592455392, "elapsed_seconds": 14109.916173283942, "env_steps_per_second": 4110.58430735532, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4866769914379632, "eta_seconds": 5352.037169176668, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0006770159199660508, "chaser_grad_norm": 0.04349134210497141, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 64.44725799560547, "collisions_per_episode": 1.0500000715255737, "entropy": 0.00310589128639549, "explorer_grad_norm": 0.5458345338702202, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 113.97291564941406, "final_distance": 2.190478563308716, "max_param_abs": 2.4299399852752686, "policy_loss": -0.00025005317962478735, "value_loss": 0.11118910647928715}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:53:18.091056+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036398.0910678, "update": 14500}
+{"device": "cuda", "elapsed_hours": 4.054119260596118, "elapsed_seconds": 14594.829338146024, "env_steps_per_second": 4111.044987910888, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3513730868653726, "eta_seconds": 4864.943112715341, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.002014808165455406, "chaser_grad_norm": 0.09840812161564827, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 64.50088500976562, "collisions_per_episode": 1.0, "entropy": 0.003318981733173132, "explorer_grad_norm": 0.2116052582859993, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 119.79007720947266, "final_distance": 2.0650737285614014, "max_param_abs": 2.4542784690856934, "policy_loss": -0.0007690534833804108, "value_loss": 0.1076736319810152}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:01:23.004223+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036883.004239, "update": 15000}
+{"device": "cuda", "elapsed_hours": 4.192042195535, "elapsed_seconds": 15091.351903925999, "env_steps_per_second": 4108.313184577637, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2170445083811288, "eta_seconds": 4381.360230172064, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 3.25234731755053e-05, "chaser_grad_norm": 0.0462856306694448, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.805000305175781, "chaser_subspace_norm": 65.58392333984375, "collisions_per_episode": 1.100000023841858, "entropy": 0.0020954415667802095, "explorer_grad_norm": 0.45636197179555893, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 114.12955474853516, "final_distance": 1.9388630390167236, "max_param_abs": 2.4749927520751953, "policy_loss": -0.00025378973339762467, "value_loss": 0.0852540098130703}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:09:39.526788+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037379.526802, "update": 15500}
+{"device": "cuda", "elapsed_hours": 4.329286794559448, "elapsed_seconds": 15585.432460414013, "env_steps_per_second": 4106.398726025463, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.082321698639862, "eta_seconds": 3896.3581151035037, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0016604845470581386, "chaser_grad_norm": 0.06046704109758139, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 63.56222915649414, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0028125346871092916, "explorer_grad_norm": 0.4928126856684685, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 120.77127838134766, "final_distance": 2.086970806121826, "max_param_abs": 2.497678518295288, "policy_loss": -0.0012127018691963087, "value_loss": 0.08045056462287903}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:17:53.607346+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037873.6073606, "update": 16000}
+{"device": "cuda", "elapsed_hours": 4.462656198671676, "elapsed_seconds": 16065.562315218034, "env_steps_per_second": 4108.166194561505, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9466240421424769, "eta_seconds": 3407.8465517129166, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.017760501632627504, "chaser_grad_norm": 0.05853390693664551, "chaser_new_fields": 95.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.944999694824219, "chaser_subspace_norm": 63.48599624633789, "collisions_per_episode": 0.875, "entropy": 0.002385914354817942, "explorer_grad_norm": 0.3553185537457466, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 115.29452514648438, "final_distance": 1.882149577140808, "max_param_abs": 2.5180764198303223, "policy_loss": -9.492970397051437e-05, "value_loss": 0.09502151980996132}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:25:53.737198+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038353.737209, "update": 16500}
+{"device": "cuda", "elapsed_hours": 4.599196517355288, "elapsed_seconds": 16557.107462479034, "env_steps_per_second": 4106.997563076674, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8116229148274036, "eta_seconds": 2921.8424933786528, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0009985900814508675, "chaser_grad_norm": 0.08174319146201015, "chaser_new_fields": 95.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.940000534057617, "chaser_subspace_norm": 64.16216278076172, "collisions_per_episode": 0.699999988079071, "entropy": 0.003951519611291587, "explorer_grad_norm": 0.40092281624674797, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 122.30136108398438, "final_distance": 1.778633952140808, "max_param_abs": 2.5919830799102783, "policy_loss": -0.00033798273565466275, "value_loss": 0.053136701695621014}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:05.282348+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038845.2823637, "update": 17000}
+{"device": "cuda", "elapsed_hours": 4.73444926007751, "elapsed_seconds": 17044.017336279037, "env_steps_per_second": 4107.012954686541, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6763498942967873, "eta_seconds": 2434.859619468434, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0001943756072434466, "chaser_grad_norm": 0.039517590776085854, "chaser_new_fields": 95.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.9350004196167, "chaser_subspace_norm": 69.97343444824219, "collisions_per_episode": 1.0500000715255737, "entropy": 0.001038317845086567, "explorer_grad_norm": 0.566088929772377, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.6875, "explorer_subspace_norm": 118.8498764038086, "final_distance": 1.9658924341201782, "max_param_abs": 2.7334179878234863, "policy_loss": -0.00018318272277229042, "value_loss": 0.07265921123325825}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:12.192219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039332.1922312, "update": 17500}
+{"device": "cuda", "elapsed_hours": 4.8722572013183445, "elapsed_seconds": 17540.12592474604, "env_steps_per_second": 4104.8736085993905, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5413619112575938, "eta_seconds": 1948.9028805273379, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00040952912746639036, "chaser_grad_norm": 0.06409427057951689, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 68.85417175292969, "collisions_per_episode": 1.0, "entropy": 0.002259656845126301, "explorer_grad_norm": 0.3050736915320158, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 126.66679382324219, "final_distance": 1.6908377408981323, "max_param_abs": 2.8545033931732178, "policy_loss": -0.00034200622106395073, "value_loss": 0.11007564514875412}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:50:28.300808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039828.3008223, "update": 18000}
+{"device": "cuda", "elapsed_hours": 5.005594914789447, "elapsed_seconds": 18020.14169324201, "env_steps_per_second": 4106.515989702333, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40585904714509036, "eta_seconds": 1461.0925697223254, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0029222532060817485, "chaser_grad_norm": 0.10623816214501858, "chaser_new_fields": 94.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 70.37963104248047, "collisions_per_episode": 1.0500000715255737, "entropy": 0.002306440786924213, "explorer_grad_norm": 0.5238702520728111, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 122.44673919677734, "final_distance": 1.840972900390625, "max_param_abs": 2.980407953262329, "policy_loss": -0.0007874917932753078, "value_loss": 0.14899397641420364}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:58:28.316576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040308.3165872, "update": 18500}
+{"device": "cuda", "elapsed_hours": 5.1398212427180585, "elapsed_seconds": 18503.35647378501, "env_steps_per_second": 4107.362905085598, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2705169075114767, "eta_seconds": 973.8608670413163, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0019687057694612875, "chaser_grad_norm": 0.11068585515022278, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 69.50489807128906, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004220188595354557, "explorer_grad_norm": 0.8308986872434616, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 132.56944274902344, "final_distance": 2.0065629482269287, "max_param_abs": 3.1132256984710693, "policy_loss": -0.0011765843178110913, "value_loss": 0.10864873044192791}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:06:31.531358+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040791.531373, "update": 19000}
+{"device": "cuda", "elapsed_hours": 5.273368907361938, "elapsed_seconds": 18984.128066502977, "env_steps_per_second": 4108.695417917511, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13521458736825484, "eta_seconds": 486.7725145257174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00016270278459473708, "chaser_grad_norm": 0.04523493256419897, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 71.9196548461914, "collisions_per_episode": 0.800000011920929, "entropy": 0.0027996512362733483, "explorer_grad_norm": 0.46223966032266617, "explorer_new_fields": 97.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 128.7664031982422, "final_distance": 2.235522985458374, "max_param_abs": 3.252021074295044, "policy_loss": -0.00030302145377447687, "value_loss": 0.08791894093155861}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:14:32.302950+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041272.3029628, "update": 19500}
+{"device": "cuda", "elapsed_hours": 5.406687021900014, "elapsed_seconds": 19464.07327884005, "env_steps_per_second": 4110.136601621321, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:22:32.248162+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041752.2481732, "update": 20000}
+{"device": "cuda", "elapsed_hours": 5.406713302494997, "elapsed_seconds": 19464.16788898199, "env_steps_per_second": 4110.116623340745, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:22:32.342773+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041752.3427885, "update": 20000}

non_social/seed_0006/attempt_01/status.json

@@ -0,0 +1,40 @@
+{
+  "device": "cuda",
+  "elapsed_hours": 5.406713302494997,
+  "elapsed_seconds": 19464.16788898199,
+  "env_steps_per_second": 4110.116623340745,
+  "environment_steps": 80000000,
+  "error": null,
+  "estimated_cost_usd": null,
+  "eta_hours": 0.0,
+  "eta_seconds": 0.0,
+  "healthy": true,
+  "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors",
+  "metrics": {
+    "approx_kl": 5.701445683692352e-05,
+    "chaser_grad_norm": 0.033485319931060076,
+    "chaser_new_fields": 94.57500457763672,
+    "chaser_partner_vision": 0.0,
+    "chaser_return": 8.830000877380371,
+    "chaser_subspace_norm": 70.88212585449219,
+    "collisions_per_episode": 0.9000000357627869,
+    "entropy": 0.0015748643199913204,
+    "explorer_grad_norm": 0.313498567789793,
+    "explorer_new_fields": 97.55000305175781,
+    "explorer_partner_vision": 0.0,
+    "explorer_return": 95.80000305175781,
+    "explorer_subspace_norm": 132.86404418945312,
+    "final_distance": 2.430217742919922,
+    "max_param_abs": 3.379441976547241,
+    "policy_loss": -4.615307129274271e-05,
+    "value_loss": 0.07251793146133423
+  },
+  "progress_fraction": 1.0,
+  "schema_version": 2,
+  "state": "completed",
+  "timestamp": "2026-07-03T01:22:32.342773+00:00",
+  "total_environment_steps": 80000000,
+  "total_updates": 20000,
+  "unix_time": 1783041752.3427885,
+  "update": 20000
+}