diff --git a/masters/abc-sim/ball_tray/meta.json b/masters/abc-sim/ball_tray/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..3abad015f0e442388c00d2d28b66135c10efc889 --- /dev/null +++ b/masters/abc-sim/ball_tray/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "ball_tray", + "task_dir": "abc-sim-ball-tray-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-ball-tray-i00-privileged/abc-sim-ball-tray-i00-privileged__BpBt4SA", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-ball-tray-i00-privileged/abc-sim-ball-tray-i00-privileged__BpBt4SA", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 551, + "seconds": 18.37, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps; ball tray: the countdown completed with the ball on the tray and both handles held), the end-state hash equals both grading replays' hash (22dc6e5abac334f2), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "tray", + "reason": "success", + "steps": 480, + "success_run": 0, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 480, + "control_steps_replayed": 480, + "sim_time_s": 16.32, + "frames_every_control_steps": 0.9804, + "final_state_sha": "22dc6e5abac334f2", + "plain_replay_state_sha": "22dc6e5abac334f2", + "replays_agree": true, + "instance_sha": "6cdaa86a450092e7", + "init_sha": "6cdaa86a450092e7", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "a5043076c687d194e7a0aea36610fd08dfbe9de32550c62449dd74e0f265ef02", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "22dc6e5abac334f2" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4355, + -0.0149, + 0.8922 + ], + "distance_m": 1.6022, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.65, + "created": "2026-10-31T05:25:26-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/ball_tray/poster.png b/masters/abc-sim/ball_tray/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..275ff1567b3e8becd095b7ba7dd2ea8535ada44a --- /dev/null +++ b/masters/abc-sim/ball_tray/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:673e2ffeb9554e37be32b20cbe3fd1cefb119ecccca8904db314428ee884af81 +size 3250569 diff --git a/masters/abc-sim/ball_tray/video.mp4 b/masters/abc-sim/ball_tray/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f148f8f5545cb71639d446e7585e54717381a321 --- /dev/null +++ b/masters/abc-sim/ball_tray/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:915da35c5ea292015d09be089871d3946b6b9a6d2f93f528cd9efd381c4affb7 +size 9130694 diff --git a/masters/abc-sim/chess/meta.json b/masters/abc-sim/chess/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..771a414d8d424ace55e716224b972d8d53075ac4 --- /dev/null +++ b/masters/abc-sim/chess/meta.json @@ -0,0 +1,79 @@ +{ + "suite": "abc-sim", + "task": "chess", + "task_dir": "abc-sim-chess-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-chess-i00-privileged/abc-sim-chess-i00-privileged__C6wgrLQ", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-chess-i00-privileged/abc-sim-chess-i00-privileged__C6wgrLQ", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 699, + "seconds": 23.3, + "speed": 12.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (4efa1720e99e91e1), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 7495, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 7496, + "control_steps_replayed": 7495, + "sim_time_s": 254.83, + "frames_every_control_steps": 11.7647, + "final_state_sha": "4efa1720e99e91e1", + "plain_replay_state_sha": "4efa1720e99e91e1", + "replays_agree": true, + "instance_sha": "65f14022364f054f", + "init_sha": "65f14022364f054f", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "4f75c780aee4c406e4e0372fbdb4ee7097bf2b46010afe6020042c84cb563510", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "4efa1720e99e91e1" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)", + "unused objects the scene parks far below or above the table (they fall freely and touch nothing) not drawn: tin_box" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5429, + -0.0119, + 0.8681 + ], + "distance_m": 1.8784, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.8, + "created": "2026-10-31T04:41:08-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/chess/poster.png b/masters/abc-sim/chess/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..ef70b24eab1f4084671de80910cf413f980bd591 --- /dev/null +++ b/masters/abc-sim/chess/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f79b68addeff799eb329ae4734a91e5a5bb9ce0452ec92f8a067384fb28e855f +size 3298051 diff --git a/masters/abc-sim/chess/video.mp4 b/masters/abc-sim/chess/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3667cc24a08d592a5480943fe15166d4063b17ee --- /dev/null +++ b/masters/abc-sim/chess/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c5be7160e8e7b79400aca103b19286cbb0641fe604ce5c1d2954f728351f0baf +size 12097385 diff --git a/masters/abc-sim/conveyor_pick/meta.json b/masters/abc-sim/conveyor_pick/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..b6a7bf42559bd475680f6beb2bec723ebfb2c9ff --- /dev/null +++ b/masters/abc-sim/conveyor_pick/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "conveyor_pick", + "task_dir": "abc-sim-conveyor-pick-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-conveyor-pick-i00-privileged/abc-sim-conveyor-pick-i00-privil__gMbxpq7", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-conveyor-pick-i00-privileged/abc-sim-conveyor-pick-i00-privil__gMbxpq7", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 391, + "seconds": 13.03, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (58f73560a19e6d6f), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 324, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 324, + "control_steps_replayed": 324, + "sim_time_s": 11.016, + "frames_every_control_steps": 0.9804, + "final_state_sha": "58f73560a19e6d6f", + "plain_replay_state_sha": "58f73560a19e6d6f", + "replays_agree": true, + "instance_sha": "73fd8e08bd4000ee", + "init_sha": "73fd8e08bd4000ee", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "e6cd92bde92818e7540e6885adcfc9fad598115f51c1852a0b4e768f7e549774", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "58f73560a19e6d6f" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.414, + 0.0774, + 0.8928 + ], + "distance_m": 1.6922, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 2.87, + "created": "2026-10-31T05:25:29-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/conveyor_pick/poster.png b/masters/abc-sim/conveyor_pick/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..da49c9b676bd046a11481bcb4439bfa87a4f4cdb --- /dev/null +++ b/masters/abc-sim/conveyor_pick/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:903c811ffe28a14b3d2224a73faf9cdda730509407fd583cf9a08b068b0ce76c +size 3428234 diff --git a/masters/abc-sim/conveyor_pick/video.mp4 b/masters/abc-sim/conveyor_pick/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4f419dae5eb50fd745fe448207f163eb32d79ad6 --- /dev/null +++ b/masters/abc-sim/conveyor_pick/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66a6ba5c549ccf870b04f2d20918a6372729252e07a0a38150cb420f58c84356 +size 8474381 diff --git a/masters/abc-sim/count/meta.json b/masters/abc-sim/count/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..f708117c869f48cd84df67186caa94d9abea7c4d --- /dev/null +++ b/masters/abc-sim/count/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "count", + "task_dir": "abc-sim-count-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-count-i00-privileged/abc-sim-count-i00-privileged__AM5P4XK", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-count-i00-privileged/abc-sim-count-i00-privileged__AM5P4XK", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 564, + "seconds": 18.8, + "speed": 4.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (04b33d53e5808f8a), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 1972, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 1972, + "control_steps_replayed": 1972, + "sim_time_s": 67.048, + "frames_every_control_steps": 3.9216, + "final_state_sha": "04b33d53e5808f8a", + "plain_replay_state_sha": "04b33d53e5808f8a", + "replays_agree": true, + "instance_sha": "080ddb7334cfb5d4", + "init_sha": "080ddb7334cfb5d4", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "6bde4f097a82747a243b1141ab2d631d1b0e1161e54748f6fda5ba1f8edba17c", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "04b33d53e5808f8a" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4595, + 0.0087, + 0.9239 + ], + "distance_m": 1.6569, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.76, + "created": "2026-10-31T06:36:22-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/count/poster.png b/masters/abc-sim/count/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..55561e10c66380595a769743e76db40e3c7a0dfe --- /dev/null +++ b/masters/abc-sim/count/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:473fb364be8c9e9a33474e08f2909b40937e462e283b94f200c32bc59345679a +size 3290066 diff --git a/masters/abc-sim/count/video.mp4 b/masters/abc-sim/count/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c4c2c507dab349163dccd5ac45948cb8c6773bf4 --- /dev/null +++ b/masters/abc-sim/count/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4ad490bd40cc30dcf9d41553320d23ceafcc1e36230262ec8297ed1c4367246 +size 8719049 diff --git a/masters/abc-sim/count_four/meta.json b/masters/abc-sim/count_four/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..57f54853e816ec188a5518dd43e794157f32cd58 --- /dev/null +++ b/masters/abc-sim/count_four/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "count_four", + "task_dir": "abc-sim-count-four-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-count-four-i00-privileged/abc-sim-count-four-i00-privilege__oMffyCC", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-count-four-i00-privileged/abc-sim-count-four-i00-privilege__oMffyCC", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 648, + "seconds": 21.6, + "speed": 8.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (5a16271694057122), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 4604, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 4604, + "control_steps_replayed": 4604, + "sim_time_s": 156.536, + "frames_every_control_steps": 7.8431, + "final_state_sha": "5a16271694057122", + "plain_replay_state_sha": "5a16271694057122", + "replays_agree": true, + "instance_sha": "c8a0c7b3eb8622b4", + "init_sha": "c8a0c7b3eb8622b4", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "62358acbf400ad4be5b0324a7ac799c94b97d07f123ecd121ecfa47dc9ad77a3", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "5a16271694057122" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4584, + 0.0256, + 1.02 + ], + "distance_m": 1.9123, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 10.18, + "created": "2026-10-31T06:47:48-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/count_four/poster.png b/masters/abc-sim/count_four/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..5c76e52972b651888922f84ea72e928e3406868e --- /dev/null +++ b/masters/abc-sim/count_four/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8701b0d23295dcb0fe6be9f634fb1b9731cdae60306eb3e7a5360489b60cc8b +size 3185665 diff --git a/masters/abc-sim/count_four/video.mp4 b/masters/abc-sim/count_four/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..2958cc4bb0884bace3f72953c59a87eea696fb08 --- /dev/null +++ b/masters/abc-sim/count_four/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95f5c55778143c05d609d8f0d6c300ec779ae5ee9aacc46674a040f0b18beb4c +size 9420711 diff --git a/masters/abc-sim/count_one/meta.json b/masters/abc-sim/count_one/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..14e496d898fc499386b578804c3f0fdcf1660225 --- /dev/null +++ b/masters/abc-sim/count_one/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "count_one", + "task_dir": "abc-sim-count-one-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-count-one-i00-privileged/abc-sim-count-one-i00-privileged__4hBS3vt", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-count-one-i00-privileged/abc-sim-count-one-i00-privileged__4hBS3vt", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 212, + "seconds": 7.07, + "speed": 0.5, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (21a0e0be11ba1e73), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 74, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 74, + "control_steps_replayed": 74, + "sim_time_s": 2.516, + "frames_every_control_steps": 0.4902, + "final_state_sha": "21a0e0be11ba1e73", + "plain_replay_state_sha": "21a0e0be11ba1e73", + "replays_agree": true, + "instance_sha": "39fcd07ad171ba75", + "init_sha": "39fcd07ad171ba75", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "9dae15aa7935b47d5fe0d2097df5f201f78c979179f7a9e20059df6aef95970c", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "21a0e0be11ba1e73" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.458, + 0.0434, + 0.9225 + ], + "distance_m": 1.5963, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.52, + "created": "2026-10-31T06:02:03-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/count_one/poster.png b/masters/abc-sim/count_one/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..70d9a2fa70b7d1b911240c89014472ed29f9f65f --- /dev/null +++ b/masters/abc-sim/count_one/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c710c4929556c9f10b4d7df62805df4d16180a74e4bb1850ad0de51a99720d39 +size 3286777 diff --git a/masters/abc-sim/count_one/video.mp4 b/masters/abc-sim/count_one/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..374e1f793ff7c91655570880f6d36e8ec20bf51b --- /dev/null +++ b/masters/abc-sim/count_one/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:81df79ac63293b45ed2bbd1a598061cc619dc22ff2a087f24a9e6c6086d79e18 +size 3469956 diff --git a/masters/abc-sim/count_three/meta.json b/masters/abc-sim/count_three/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..151a97acecdf08cb148206d4c1c76098ff3d0281 --- /dev/null +++ b/masters/abc-sim/count_three/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "count_three", + "task_dir": "abc-sim-count-three-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-count-three-i00-privileged/abc-sim-count-three-i00-privileg__LgzsQAJ", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-count-three-i00-privileged/abc-sim-count-three-i00-privileg__LgzsQAJ", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 647, + "seconds": 21.57, + "speed": 1.5, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (16cde7e8423c0ce3), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 862, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 862, + "control_steps_replayed": 862, + "sim_time_s": 29.308, + "frames_every_control_steps": 1.4706, + "final_state_sha": "16cde7e8423c0ce3", + "plain_replay_state_sha": "16cde7e8423c0ce3", + "replays_agree": true, + "instance_sha": "4a649f7e73ba4df1", + "init_sha": "4a649f7e73ba4df1", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "d3648bd3fef062b205af60be9812dfe5d3dae3662cb73c308216ee91421acba5", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "16cde7e8423c0ce3" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4264, + 0.0576, + 0.8823 + ], + "distance_m": 1.5317, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 2.84, + "created": "2026-10-31T04:41:13-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/count_three/poster.png b/masters/abc-sim/count_three/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..4c814774aa71a96d5ce3dd4b16670e9c4e8fdfbc --- /dev/null +++ b/masters/abc-sim/count_three/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b99bb1e71ed5b5caf5706afe73e7f56081cfb23952047cb2d8d77ac8162d4ac2 +size 3262582 diff --git a/masters/abc-sim/count_three/video.mp4 b/masters/abc-sim/count_three/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..acbfb6522252a43730f4cadc2afff561c46d36a8 --- /dev/null +++ b/masters/abc-sim/count_three/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:abdb7e6c2007d0cafc85c4977ee6973d83d2634a55c7c415a8d4dd63e328e683 +size 10685828 diff --git a/masters/abc-sim/count_two/meta.json b/masters/abc-sim/count_two/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..cbd2f72c0d52631cf24430ee7e938e10ce16c704 --- /dev/null +++ b/masters/abc-sim/count_two/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "count_two", + "task_dir": "abc-sim-count-two-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-count-two-i00-privileged/abc-sim-count-two-i00-privileged__hjJa4h2", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-count-two-i00-privileged/abc-sim-count-two-i00-privileged__hjJa4h2", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 448, + "seconds": 14.93, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (1d0fc8f13afd69fa), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 379, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 379, + "control_steps_replayed": 379, + "sim_time_s": 12.886, + "frames_every_control_steps": 0.9804, + "final_state_sha": "1d0fc8f13afd69fa", + "plain_replay_state_sha": "1d0fc8f13afd69fa", + "replays_agree": true, + "instance_sha": "5099066e29874d64", + "init_sha": "5099066e29874d64", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "535dcc951ef08cddca38241fedd73d9634446ddb98c461207c0f59aab4532cfe", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "1d0fc8f13afd69fa" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.452, + 0.0334, + 0.9174 + ], + "distance_m": 1.6136, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 3.04, + "created": "2026-10-31T05:38:11-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/count_two/poster.png b/masters/abc-sim/count_two/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..28e9cb9b3d3141d4c3605fa05ca8b821c77c7e53 --- /dev/null +++ b/masters/abc-sim/count_two/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ef9b77dbba4c8d99d27b989c636633aef870707a13fe6e7cd3d189fd0bef01c +size 3271462 diff --git a/masters/abc-sim/count_two/video.mp4 b/masters/abc-sim/count_two/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..352d27e95109079f3662c52e1de13cc6cc96a054 --- /dev/null +++ b/masters/abc-sim/count_two/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90d9100bf161041edd2e3b6c6450e53358849405b04711dcc8967f30c6ef7055 +size 7076838 diff --git a/masters/abc-sim/dish_rack/meta.json b/masters/abc-sim/dish_rack/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..ab3264a15fafc856dff10c03a760e710f4d2d171 --- /dev/null +++ b/masters/abc-sim/dish_rack/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "dish_rack", + "task_dir": "abc-sim-dish-rack-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-dish-rack-i00-privileged/abc-sim-dish-rack-i00-privileged__aKL9b6V", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-dish-rack-i00-privileged/abc-sim-dish-rack-i00-privileged__aKL9b6V", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 648, + "seconds": 21.6, + "speed": 1.5, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (e94b0ef819675914), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 863, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 863, + "control_steps_replayed": 863, + "sim_time_s": 29.342, + "frames_every_control_steps": 1.4706, + "final_state_sha": "e94b0ef819675914", + "plain_replay_state_sha": "e94b0ef819675914", + "replays_agree": true, + "instance_sha": "f008c6c643f75ca1", + "init_sha": "f008c6c643f75ca1", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "b48fabf316c93acd20b4479261f27409f45a01e52db6ded514450c8015454406", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "e94b0ef819675914" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4782, + 0.0163, + 0.9157 + ], + "distance_m": 1.51, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.41, + "created": "2026-10-31T05:38:15-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/dish_rack/poster.png b/masters/abc-sim/dish_rack/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..281a30e3650d66777f0f36f4b80e8ec3596c48cb --- /dev/null +++ b/masters/abc-sim/dish_rack/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57bc8656eb574514f382b3b6fa7981e09ff3a6c1605773a054dd4f84d2b79e50 +size 3360624 diff --git a/masters/abc-sim/dish_rack/video.mp4 b/masters/abc-sim/dish_rack/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5908ff8d4ba711a0693ec5f1326afd0937a91caf --- /dev/null +++ b/masters/abc-sim/dish_rack/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c904215b916baea14c610f8c0e0fbb4521f006f13aaed03dca5213f8282e536 +size 12167666 diff --git a/masters/abc-sim/drawer_search/meta.json b/masters/abc-sim/drawer_search/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..785f4bea6e3431538b26ff991c92f3bfc98e280a --- /dev/null +++ b/masters/abc-sim/drawer_search/meta.json @@ -0,0 +1,80 @@ +{ + "suite": "abc-sim", + "task": "drawer_search", + "task_dir": "abc-sim-drawer-search-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-drawer-search-i00-privileged/abc-sim-drawer-search-i00-privil__anJwWjZ", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-drawer-search-i00-privileged/abc-sim-drawer-search-i00-privil__anJwWjZ", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 682, + "seconds": 22.73, + "speed": 6.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (c774fb4ae8abfd70), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 3653, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 3653, + "control_steps_replayed": 3653, + "sim_time_s": 124.202, + "frames_every_control_steps": 5.8824, + "final_state_sha": "c774fb4ae8abfd70", + "plain_replay_state_sha": "c774fb4ae8abfd70", + "replays_agree": true, + "instance_sha": "c2f19b21de544dd2", + "init_sha": "c2f19b21de544dd2", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "b29470b6af4515c5b3a50bea07d69d118a5b5122d0966373950987fb09a03c85", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "c774fb4ae8abfd70" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)", + "unused objects the scene parks far below or above the table (they fall freely and touch nothing) not drawn: drawer_search_distractor_01, drawer_search_distractor_02, drawer_search_distractor_03, drawer_search_distractor_04, drawer_search_distractor_05, drawer_search_distractor_06, drawer_search_distractor_07, drawer_search_distractor_08, drawer_search_distractor_09, drawer_search_distractor_10, drawer_search_distractor_11, drawer_search_distractor_13, drawer_search_distractor_16, drawer_search_distractor_18, drawer_search_distractor_19, drawer_search_distractor_20, drawer_search_distractor_21, drawer_search_distractor_22, drawer_search_distractor_23, drawer_search_distractor_24, drawer_search_distractor_25, drawer_search_target, drawer_stack_1, drawer_stack_2" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 155.0, + "el_deg": 40.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4354, + -0.0612, + 1.0015 + ], + "distance_m": 1.9168, + "orbit_deg": 10.0, + "note": "viewed from behind the arms: from the usual 3/4 front view the drawer stack hides the arms' work" + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 9.77, + "created": "2026-10-31T06:49:04-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/drawer_search/poster.png b/masters/abc-sim/drawer_search/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..ca7414758c135df4db29141e7ca5ae919bd9556a --- /dev/null +++ b/masters/abc-sim/drawer_search/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:478e439aa333f69d629da883e1405e51fa961d4669104eb7a8e7dfe2c913e123 +size 3193705 diff --git a/masters/abc-sim/drawer_search/video.mp4 b/masters/abc-sim/drawer_search/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d2f678cdd50605f8f83700e23ca619d74d647900 --- /dev/null +++ b/masters/abc-sim/drawer_search/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c403e25f68a5f07eec53878dc90f5eac60baa3698272be3a1c1719d81098e2a4 +size 9123286 diff --git a/masters/abc-sim/grab_clutter/meta.json b/masters/abc-sim/grab_clutter/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..00b84a0ee9a1e85f38c318415bbc07d502807fee --- /dev/null +++ b/masters/abc-sim/grab_clutter/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "grab_clutter", + "task_dir": "abc-sim-grab-clutter-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-grab-clutter-i00-privileged/abc-sim-grab-clutter-i00-privile__gcbkAVs", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-grab-clutter-i00-privileged/abc-sim-grab-clutter-i00-privile__gcbkAVs", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 425, + "seconds": 14.17, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (a51263df9f4d3c28), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 357, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 357, + "control_steps_replayed": 357, + "sim_time_s": 12.138, + "frames_every_control_steps": 0.9804, + "final_state_sha": "a51263df9f4d3c28", + "plain_replay_state_sha": "a51263df9f4d3c28", + "replays_agree": true, + "instance_sha": "a2e0c4bffcd832d7", + "init_sha": "a2e0c4bffcd832d7", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "bab6ad70e1b3ac7a85ccb8570fcee579f58858749cd843408e5491d15f7c3b65", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "a51263df9f4d3c28" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5113, + -0.0571, + 0.8954 + ], + "distance_m": 1.7274, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.63, + "created": "2026-10-31T06:47:44-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/grab_clutter/poster.png b/masters/abc-sim/grab_clutter/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..aa4891dff0568002564ea534814b0d4fbf535c0a --- /dev/null +++ b/masters/abc-sim/grab_clutter/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdd011fae4c6c9a61e108363a45746774f021c60ad8105531374bcd994bff4da +size 3292282 diff --git a/masters/abc-sim/grab_clutter/video.mp4 b/masters/abc-sim/grab_clutter/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f909cf707f392e9eef075204190b0c725992d564 --- /dev/null +++ b/masters/abc-sim/grab_clutter/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f87aa2e881d702e6e395cbde182f36bb76e61bef7537fed732b168bbe8c91019 +size 6726898 diff --git a/masters/abc-sim/hang_mug/meta.json b/masters/abc-sim/hang_mug/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..0ec4846a33e9f8a22566cae7cf3d9b1cd3bd2eeb --- /dev/null +++ b/masters/abc-sim/hang_mug/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "hang_mug", + "task_dir": "abc-sim-hang-mug-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-hang-mug-i00-privileged/abc-sim-hang-mug-i00-privileged__eEKwPb7", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-hang-mug-i00-privileged/abc-sim-hang-mug-i00-privileged__eEKwPb7", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 609, + "seconds": 20.3, + "speed": 1.5, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (59e68561c05dcde6), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 806, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 806, + "control_steps_replayed": 806, + "sim_time_s": 27.404, + "frames_every_control_steps": 1.4706, + "final_state_sha": "59e68561c05dcde6", + "plain_replay_state_sha": "59e68561c05dcde6", + "replays_agree": true, + "instance_sha": "f971caba9bc2c81c", + "init_sha": "f971caba9bc2c81c", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "13854835a1ace62a05cdd7fe130b118676d185a2491d4d898cc17f7519facded", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "59e68561c05dcde6" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4312, + 0.0217, + 0.9232 + ], + "distance_m": 1.465, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.52, + "created": "2026-10-31T04:29:10-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/hang_mug/poster.png b/masters/abc-sim/hang_mug/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..a53dbbc2278cdc44da96ca819f4159b800131bfb --- /dev/null +++ b/masters/abc-sim/hang_mug/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d5b8a145674c8bcb50ca4641f658fe35bcc2be3b296462199d04d049e4061bc +size 3302733 diff --git a/masters/abc-sim/hang_mug/video.mp4 b/masters/abc-sim/hang_mug/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3e8a8ea229c81c1053d40a3f53a51845e6ecbc40 --- /dev/null +++ b/masters/abc-sim/hang_mug/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fa4180e2e79bf2a6ddc22559fc7ae05f064ba783021c4cdb929265011a18c3b +size 10073096 diff --git a/masters/abc-sim/highlights.json b/masters/abc-sim/highlights.json new file mode 100644 index 0000000000000000000000000000000000000000..1fd01d25c2dfa9bd309d34dc0d86a095a2084c18 --- /dev/null +++ b/masters/abc-sim/highlights.json @@ -0,0 +1,20 @@ +{ + "teaser": "lego", + "highlights": [ + "lego", + "nuts_bolts", + "hang_mug", + "chess", + "throw_bottles", + "pouring" + ], + "why": { + "lego": "both arms sort colour bricks into two bins, then one bin is tipped into the tote", + "nuts_bolts": "the hardest task: no GPT-6 Luna success in any mode; GPT-6 Astra's standard success, both arms sorting nuts and bolts", + "hang_mug": "a mug handed from one arm to the other and hung on the tree", + "chess": "pieces picked off the board and placed with both arms", + "throw_bottles": "bottles picked up and dropped into the bin", + "pouring": "pouring from one cup into the other" + }, + "why_fewer": "" +} \ No newline at end of file diff --git a/masters/abc-sim/inhand_transfer/meta.json b/masters/abc-sim/inhand_transfer/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..d301e63d1fee938909021c65ca8e82d1ceca0b12 --- /dev/null +++ b/masters/abc-sim/inhand_transfer/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "inhand_transfer", + "task_dir": "abc-sim-inhand-transfer-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-inhand-transfer-i00-privileged/abc-sim-inhand-transfer-i00-priv__bB2rjgH", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-inhand-transfer-i00-privileged/abc-sim-inhand-transfer-i00-priv__bB2rjgH", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 285, + "seconds": 9.5, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (3ac4b107f71cbcb8), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 220, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 220, + "control_steps_replayed": 220, + "sim_time_s": 7.48, + "frames_every_control_steps": 0.9804, + "final_state_sha": "3ac4b107f71cbcb8", + "plain_replay_state_sha": "3ac4b107f71cbcb8", + "replays_agree": true, + "instance_sha": "2fee35e343f1e9d5", + "init_sha": "2fee35e343f1e9d5", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "fd5659868b6c420f89dbb0d3454dd6da3b10a1edcf8895b0bfbc016c31772043", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "3ac4b107f71cbcb8" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4653, + 0.0452, + 0.9364 + ], + "distance_m": 1.4135, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.4, + "created": "2026-10-31T06:15:20-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/inhand_transfer/poster.png b/masters/abc-sim/inhand_transfer/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..d42c6e3f9d11be7ebd59ed030a3a4bba9e437342 --- /dev/null +++ b/masters/abc-sim/inhand_transfer/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:21552b2b66e2c876ddd3145fbec94a669fa2d389410d210c68ee117053d02132 +size 3206396 diff --git a/masters/abc-sim/inhand_transfer/video.mp4 b/masters/abc-sim/inhand_transfer/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a53d7618d1e0ba8a7d9bf34dca7deab59c642c48 --- /dev/null +++ b/masters/abc-sim/inhand_transfer/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b97d9b9772f8ad840b744dff686c68005f18e54e2124c0ba09f35b84e5f1c5c6 +size 5183804 diff --git a/masters/abc-sim/lego/meta.json b/masters/abc-sim/lego/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..8a2279baa50870a5336a5f3f1e99c2fb69cd63b7 --- /dev/null +++ b/masters/abc-sim/lego/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "lego", + "task_dir": "abc-sim-lego-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-lego-i00-privileged/abc-sim-lego-i00-privileged__ku7HjL5", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-lego-i00-privileged/abc-sim-lego-i00-privileged__ku7HjL5", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 589, + "seconds": 19.63, + "speed": 20.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (96b18189a5afadc2), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 10334, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 10334, + "control_steps_replayed": 10334, + "sim_time_s": 351.356, + "frames_every_control_steps": 19.6078, + "final_state_sha": "96b18189a5afadc2", + "plain_replay_state_sha": "96b18189a5afadc2", + "replays_agree": true, + "instance_sha": "2ebc48bf64974f98", + "init_sha": "2ebc48bf64974f98", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "3974f80dec008fb9855d149fd58b32bba9ae9629b1fe41e702c4cb966998267f", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "96b18189a5afadc2" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4871, + 0.0341, + 0.9031 + ], + "distance_m": 1.7469, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 2.93, + "created": "2026-10-31T04:08:17-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/lego/poster.png b/masters/abc-sim/lego/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..2b6cd2e23fd8353b21cd1f22b401ec2efec4b5d1 --- /dev/null +++ b/masters/abc-sim/lego/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:370911b4da00e4baaeda83f07d69856f12d4ec44864527df2a4176f95c8d9517 +size 3237368 diff --git a/masters/abc-sim/lego/video.mp4 b/masters/abc-sim/lego/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4341711bc9092a878dc9597aaa3da230118a2676 --- /dev/null +++ b/masters/abc-sim/lego/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd0a43fb19da550869df6db66138c5bfc3c132e8babf8d845989a660948c6f10 +size 9892185 diff --git a/masters/abc-sim/nuts_bolts/meta.json b/masters/abc-sim/nuts_bolts/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..eab3504d083180d4f82cd88836db55e08b9c0e88 --- /dev/null +++ b/masters/abc-sim/nuts_bolts/meta.json @@ -0,0 +1,79 @@ +{ + "suite": "abc-sim", + "task": "nuts_bolts", + "task_dir": "abc-sim-nuts-bolts-i00-standard", + "source": "agent-success:codex-gpt6_astra-high-abc-sim-nuts-bolts-i00-standard/abc-sim-nuts-bolts-i00-standard__KAo3uKg", + "agent": "codex + GPT-6 Astra (high)", + "run": "codex-gpt6_astra-high", + "mode": "standard", + "trial": "codex-gpt6_astra-high-abc-sim-nuts-bolts-i00-standard/abc-sim-nuts-bolts-i00-standard__KAo3uKg", + "trial_choice": "no privileged success on this task; GPT-6 Luna failed it in every mode, and a counted success by another agent is shown before the reference solution: GPT-6 Astra's standard trial, re-rendered because its replay is bit-exact", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 713, + "seconds": 23.77, + "speed": 30.0, + "success_check": true, + "success_check_how": "standard-mode trial (graded by its live episode): re-rendered only because the replay is bit-exact. The recorded rows (the sim service's trajectory.npz) were replayed from the task's frozen instance (byte-identical in the privileged and standard directories) with the task's own tooling (rcb_abc.Session, the session the sim service runs): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals the live episode's recorded hash (539fdec941c59936), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 19156, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 19156, + "control_steps_replayed": 19156, + "sim_time_s": 651.304, + "frames_every_control_steps": 29.4118, + "final_state_sha": "539fdec941c59936", + "plain_replay_state_sha": "539fdec941c59936", + "replays_agree": true, + "instance_sha": "36a10f65fd90f73b", + "init_sha": "36a10f65fd90f73b", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "231517cc43bb082a8a0f32dc7654f5f03fa99d004a54d3dfc5c1ba4a37dae716", + "mujoco": "3.6.0", + "recorded_live_state_sha": "539fdec941c59936", + "matches_live_episode": true + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4552, + -0.0151, + 0.9698 + ], + "distance_m": 2.0078, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.2.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.55, + "created": "2026-10-31T07:22:47-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/nuts_bolts/poster.png b/masters/abc-sim/nuts_bolts/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..e029de6a45d914084119dd67c8ae812c7f8d56f1 --- /dev/null +++ b/masters/abc-sim/nuts_bolts/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92dff4bee5dbc8d9fa256e8fab2cb964204a90e13fa165b677fdf3878c4367c5 +size 3209557 diff --git a/masters/abc-sim/nuts_bolts/video.mp4 b/masters/abc-sim/nuts_bolts/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..804e4b8c6ea47d0bfeeb8379fe356ff6807d25f8 --- /dev/null +++ b/masters/abc-sim/nuts_bolts/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c384a6ded59bc4e632fd86ea8e15c557aa31e09bfd63d281a3d8da2bfc8ab2d8 +size 10853677 diff --git a/masters/abc-sim/pouring/meta.json b/masters/abc-sim/pouring/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..e1c808e0f163110a0dab615d1cb3a458767a8573 --- /dev/null +++ b/masters/abc-sim/pouring/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "pouring", + "task_dir": "abc-sim-pouring-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-pouring-i00-privileged/abc-sim-pouring-i00-privileged__oa56369", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-pouring-i00-privileged/abc-sim-pouring-i00-privileged__oa56369", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 632, + "seconds": 21.07, + "speed": 5.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (ebbb534731924b41), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 2798, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 2798, + "control_steps_replayed": 2798, + "sim_time_s": 95.132, + "frames_every_control_steps": 4.902, + "final_state_sha": "ebbb534731924b41", + "plain_replay_state_sha": "ebbb534731924b41", + "replays_agree": true, + "instance_sha": "b54e0a7dafa2a18d", + "init_sha": "b54e0a7dafa2a18d", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "16e54616c9e2a467da108ffd390297070353e526bc8bd3136daa8a26eaea4719", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "ebbb534731924b41" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4651, + 0.0504, + 0.9745 + ], + "distance_m": 1.3681, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 2.78, + "created": "2026-10-31T05:14:26-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/pouring/poster.png b/masters/abc-sim/pouring/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..f260689f106a448d4c2fe4f1d4295c69d3448b3e --- /dev/null +++ b/masters/abc-sim/pouring/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9d24a09545bb0be5d8018d0ac22822e7822b167dc52ba9330ab8a27103b98189 +size 3256159 diff --git a/masters/abc-sim/pouring/video.mp4 b/masters/abc-sim/pouring/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c4e3d4c4291a434a3781493f5f9556041425822c --- /dev/null +++ b/masters/abc-sim/pouring/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:808095474c92128cd8c86a3ec0b4f91c443d62c8e990796c03dd9eeeaf862c81 +size 10052972 diff --git a/masters/abc-sim/put_bottles/meta.json b/masters/abc-sim/put_bottles/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..d9ccf7f0beae2e8042de77b1237638a3e2b153ec --- /dev/null +++ b/masters/abc-sim/put_bottles/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "put_bottles", + "task_dir": "abc-sim-put-bottles-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-put-bottles-i00-privileged/abc-sim-put-bottles-i00-privileg__HHK8Jsa", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-put-bottles-i00-privileged/abc-sim-put-bottles-i00-privileg__HHK8Jsa", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 621, + "seconds": 20.7, + "speed": 3.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (f6857b8e253c4485), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 1647, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 1647, + "control_steps_replayed": 1647, + "sim_time_s": 55.998, + "frames_every_control_steps": 2.9412, + "final_state_sha": "f6857b8e253c4485", + "plain_replay_state_sha": "f6857b8e253c4485", + "replays_agree": true, + "instance_sha": "9caab7a4776d5af8", + "init_sha": "9caab7a4776d5af8", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "2bfeac5067246d77a1d2cb4eaf56de87aac1d06ba6611626240af22fc889f0e2", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "f6857b8e253c4485" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4334, + 0.0605, + 0.9094 + ], + "distance_m": 1.7387, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.63, + "created": "2026-10-31T05:14:30-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/put_bottles/poster.png b/masters/abc-sim/put_bottles/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..62095f55c990d0ee1d78a0985dccc12a7c3e42ba --- /dev/null +++ b/masters/abc-sim/put_bottles/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c51fd04fc7b15d5c72f944e31a58852371ebb5a42406ad3d7e801a2149c043eb +size 3238849 diff --git a/masters/abc-sim/put_bottles/video.mp4 b/masters/abc-sim/put_bottles/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..6ef3e77cf2b096db366c47b9d7a431e88a8e4aac --- /dev/null +++ b/masters/abc-sim/put_bottles/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95994fe3a8df15a8083d20e9b1e4ff1d22d8a082472f59f6bd9e5f154d145bd3 +size 10514060 diff --git a/masters/abc-sim/put_relative/meta.json b/masters/abc-sim/put_relative/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..1a33ff9d6fa16dbf8c2510a4552cbddf51c62522 --- /dev/null +++ b/masters/abc-sim/put_relative/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "put_relative", + "task_dir": "abc-sim-put-relative-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-put-relative-i00-privileged/abc-sim-put-relative-i00-privile__msmZps5", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-put-relative-i00-privileged/abc-sim-put-relative-i00-privile__msmZps5", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 278, + "seconds": 9.27, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (13d7aeaa8d157610), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 213, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 213, + "control_steps_replayed": 213, + "sim_time_s": 7.242, + "frames_every_control_steps": 0.9804, + "final_state_sha": "13d7aeaa8d157610", + "plain_replay_state_sha": "13d7aeaa8d157610", + "replays_agree": true, + "instance_sha": "f99f99ccb4ee5da8", + "init_sha": "f99f99ccb4ee5da8", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "4ebf4e3ede450cd9f77ebff1c87f8f566650b02bee3a5cd1567f0401e4703159", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "13d7aeaa8d157610" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4487, + 0.0172, + 0.9278 + ], + "distance_m": 1.4206, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.42, + "created": "2026-10-31T07:35:00-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/put_relative/poster.png b/masters/abc-sim/put_relative/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..1ec4450109c21ab8fe4c874553496d1151fa48e6 --- /dev/null +++ b/masters/abc-sim/put_relative/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c8d98a4496b9a8acd69f9f40a6e0799ce1120dddc6b2fc3a714422d0dddf6677 +size 3261211 diff --git a/masters/abc-sim/put_relative/video.mp4 b/masters/abc-sim/put_relative/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..81f2f50aaac6d4b56a220782eaf0219f000ace8e --- /dev/null +++ b/masters/abc-sim/put_relative/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d6845827830e480ee0e67a9129c8f8fc6f599a73a563eec2ce130fcf8434f1fb +size 5253222 diff --git a/masters/abc-sim/spell_abc/meta.json b/masters/abc-sim/spell_abc/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..bf0a9c89f284825694b40a7b727032bdd4ce968f --- /dev/null +++ b/masters/abc-sim/spell_abc/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_abc", + "task_dir": "abc-sim-spell-abc-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-abc-i00-privileged/abc-sim-spell-abc-i00-privileged__97DC3e4", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-abc-i00-privileged/abc-sim-spell-abc-i00-privileged__97DC3e4", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 566, + "seconds": 18.87, + "speed": 2.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (c18cb09eda224b7b), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 989, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 989, + "control_steps_replayed": 989, + "sim_time_s": 33.626, + "frames_every_control_steps": 1.9608, + "final_state_sha": "c18cb09eda224b7b", + "plain_replay_state_sha": "c18cb09eda224b7b", + "replays_agree": true, + "instance_sha": "2a83e3a237ac1ddd", + "init_sha": "2a83e3a237ac1ddd", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "182bd0005051b5d0a7ed7a0f8c174ecf2516d563cbe48fecc548a465f010f376", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "c18cb09eda224b7b" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5132, + -0.0007, + 0.897 + ], + "distance_m": 1.7949, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.68, + "created": "2026-10-31T05:38:18-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_abc/poster.png b/masters/abc-sim/spell_abc/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..2a3eeefd2dd623d7d0fc6ac86658f5dc29c84ff9 --- /dev/null +++ b/masters/abc-sim/spell_abc/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73c758e3ff4f26fe107b1e9db8943293152657cf01673f0626837bfd35b792af +size 3282445 diff --git a/masters/abc-sim/spell_abc/video.mp4 b/masters/abc-sim/spell_abc/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5d264a1f2bd39be56444a9a9261711ecf411bdf9 --- /dev/null +++ b/masters/abc-sim/spell_abc/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:669452ce4e1344fbb67620bc08a2ee6ae0ab19ad9682591909c9db8a82ab1c15 +size 9273029 diff --git a/masters/abc-sim/spell_agi/meta.json b/masters/abc-sim/spell_agi/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..4d4f39b75ceec6a37c9e644bdfda1fbfbe4c72ca --- /dev/null +++ b/masters/abc-sim/spell_agi/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_agi", + "task_dir": "abc-sim-spell-agi-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-agi-i00-privileged/abc-sim-spell-agi-i00-privileged__KSqDZdH", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-agi-i00-privileged/abc-sim-spell-agi-i00-privileged__KSqDZdH", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 624, + "seconds": 20.8, + "speed": 2.5, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (4236d3020f14391b), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 1378, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 1378, + "control_steps_replayed": 1378, + "sim_time_s": 46.852, + "frames_every_control_steps": 2.451, + "final_state_sha": "4236d3020f14391b", + "plain_replay_state_sha": "4236d3020f14391b", + "replays_agree": true, + "instance_sha": "6cc75a0bd3f48db4", + "init_sha": "6cc75a0bd3f48db4", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "8a2eb97db839fcf9bbf09ecfc4d0b744f1b70a6e3b46b35cef025ee99d2066bf", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "4236d3020f14391b" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5671, + 0.0257, + 0.8622 + ], + "distance_m": 1.7966, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.7, + "created": "2026-10-31T05:14:34-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_agi/poster.png b/masters/abc-sim/spell_agi/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..c6a177041a6d813875acf2a5a03ba5d7773a4fa0 --- /dev/null +++ b/masters/abc-sim/spell_agi/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4093d33c53963fa8624259ab97d3463b61958819011f0fa41b4d9eb818e4178 +size 3284146 diff --git a/masters/abc-sim/spell_agi/video.mp4 b/masters/abc-sim/spell_agi/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..56bf4bf5b2d581728575a0e71671aab1a00b2e2c --- /dev/null +++ b/masters/abc-sim/spell_agi/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37119681306de4c9f9d269c777ffade209555b55a33f054a8e3563861278105a +size 9393632 diff --git a/masters/abc-sim/spell_bair/meta.json b/masters/abc-sim/spell_bair/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..1978ff04a8ba8453b85243c1e6e63e76f95dadef --- /dev/null +++ b/masters/abc-sim/spell_bair/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_bair", + "task_dir": "abc-sim-spell-bair-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-bair-i00-privileged/abc-sim-spell-bair-i00-privilege__AQS6xMA", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-bair-i00-privileged/abc-sim-spell-bair-i00-privilege__AQS6xMA", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 685, + "seconds": 22.83, + "speed": 5.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (64cd136c32a32078), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 3057, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 3057, + "control_steps_replayed": 3057, + "sim_time_s": 103.938, + "frames_every_control_steps": 4.902, + "final_state_sha": "64cd136c32a32078", + "plain_replay_state_sha": "64cd136c32a32078", + "replays_agree": true, + "instance_sha": "62be37b797ce4431", + "init_sha": "62be37b797ce4431", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "9242fe46afe6aabcf240d23757714b10141ce469dcaad5deed2a03048065f290", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "64cd136c32a32078" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4704, + 0.0434, + 0.9157 + ], + "distance_m": 1.9088, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.77, + "created": "2026-10-31T04:41:18-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_bair/poster.png b/masters/abc-sim/spell_bair/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..cf370b6bcedb6b25b3099f3c53d687d7deb6a4a5 --- /dev/null +++ b/masters/abc-sim/spell_bair/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7377f87b638f02cf58fddc3b2c7d5be2913767c3fbf7af248ff9a74913cd7131 +size 3287682 diff --git a/masters/abc-sim/spell_bair/video.mp4 b/masters/abc-sim/spell_bair/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3509d8179db60c84c8c25c71756c58e1fa8d172c --- /dev/null +++ b/masters/abc-sim/spell_bair/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:953d7be46a3c66ec4337464dc49ab5e1f2ed9c85846869d3216e8bb7cef3a6e6 +size 10472975 diff --git a/masters/abc-sim/spell_cat/meta.json b/masters/abc-sim/spell_cat/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..ab0d15f4193a3bd16bca4defb03781debdedb093 --- /dev/null +++ b/masters/abc-sim/spell_cat/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_cat", + "task_dir": "abc-sim-spell-cat-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-cat-i00-privileged/abc-sim-spell-cat-i00-privileged__jdPStuR", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-cat-i00-privileged/abc-sim-spell-cat-i00-privileged__jdPStuR", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 280, + "seconds": 9.33, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (e7131f5327966efa), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 215, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 215, + "control_steps_replayed": 215, + "sim_time_s": 7.31, + "frames_every_control_steps": 0.9804, + "final_state_sha": "e7131f5327966efa", + "plain_replay_state_sha": "e7131f5327966efa", + "replays_agree": true, + "instance_sha": "8e921bf2a751b52d", + "init_sha": "8e921bf2a751b52d", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "e76504cedc6888b8989863f17f89e60e19f5e6d72bc6b221c7a1d88545bc7e90", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "e7131f5327966efa" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5495, + 0.0115, + 0.889 + ], + "distance_m": 1.6216, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.57, + "created": "2026-10-31T06:49:06-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_cat/poster.png b/masters/abc-sim/spell_cat/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..e37f48988211c18e531496d4c818e4ce841b9a76 --- /dev/null +++ b/masters/abc-sim/spell_cat/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a8ad894983db61fd5f7e5ab8043d612d945383ca4378fc00fb6668c5ead170f3 +size 3320642 diff --git a/masters/abc-sim/spell_cat/video.mp4 b/masters/abc-sim/spell_cat/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..457c86c4516d8c59cc114cb00969d0c515626438 --- /dev/null +++ b/masters/abc-sim/spell_cat/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f7b523cc89c2d91a138cbb7fa5b69d9e4f6390607a8779600b7e666e1db8fd3 +size 5518062 diff --git a/masters/abc-sim/spell_dog/meta.json b/masters/abc-sim/spell_dog/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..dc83f68387b69425e14158e67633095176c1a9c1 --- /dev/null +++ b/masters/abc-sim/spell_dog/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_dog", + "task_dir": "abc-sim-spell-dog-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-dog-i00-privileged/abc-sim-spell-dog-i00-privileged__qeb6Bdt", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-dog-i00-privileged/abc-sim-spell-dog-i00-privileged__qeb6Bdt", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 610, + "seconds": 20.33, + "speed": 4.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (8a260ce0c735a935), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 2152, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 2152, + "control_steps_replayed": 2152, + "sim_time_s": 73.168, + "frames_every_control_steps": 3.9216, + "final_state_sha": "8a260ce0c735a935", + "plain_replay_state_sha": "8a260ce0c735a935", + "replays_agree": true, + "instance_sha": "3913975861559c17", + "init_sha": "3913975861559c17", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "97aa621a12e7e4a91e90c78a2f964ea1c99909add5b2667ecb6a0aeb5bd54c09", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "8a260ce0c735a935" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5106, + 0.0794, + 0.8745 + ], + "distance_m": 1.7346, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.63, + "created": "2026-10-31T06:36:26-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_dog/poster.png b/masters/abc-sim/spell_dog/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..2ef1d6b6b48b9a395aaa9e3385d0fdcae0e3508b --- /dev/null +++ b/masters/abc-sim/spell_dog/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2e21a3dbe92dd89899dd58bd6c788263042c46f85482f3e7a17dadda9e31b812 +size 3317640 diff --git a/masters/abc-sim/spell_dog/video.mp4 b/masters/abc-sim/spell_dog/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f93697c7d4580e44770b4dc58a32987fc24618a9 --- /dev/null +++ b/masters/abc-sim/spell_dog/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6415cc03d741245cedd3d3b75ddf5f88cae79fc44c55bfb572a867b4599d800 +size 9910177 diff --git a/masters/abc-sim/spell_fish/meta.json b/masters/abc-sim/spell_fish/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..37e41c51ba6a4369ecb38b1c1126a27f11059ac1 --- /dev/null +++ b/masters/abc-sim/spell_fish/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_fish", + "task_dir": "abc-sim-spell-fish-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-fish-i00-privileged/abc-sim-spell-fish-i00-privilege__tkMFL5o", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-fish-i00-privileged/abc-sim-spell-fish-i00-privilege__tkMFL5o", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 612, + "seconds": 20.4, + "speed": 3.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (fd4d77c9975966eb), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 1620, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 1620, + "control_steps_replayed": 1620, + "sim_time_s": 55.08, + "frames_every_control_steps": 2.9412, + "final_state_sha": "fd4d77c9975966eb", + "plain_replay_state_sha": "fd4d77c9975966eb", + "replays_agree": true, + "instance_sha": "288971f6896c9bfd", + "init_sha": "288971f6896c9bfd", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "cf2bb5a490274ca3b2ddf4979849c579b6f46839b146821f44995d9f32d004dd", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "fd4d77c9975966eb" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5092, + 0.0039, + 0.8911 + ], + "distance_m": 1.7834, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.6, + "created": "2026-10-31T06:25:44-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_fish/poster.png b/masters/abc-sim/spell_fish/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..85bcb2834668cbcda3aac4583be253bffbdadf21 --- /dev/null +++ b/masters/abc-sim/spell_fish/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ebdbd46bd96c01df65761c183f7bf72b6c8406399e748c0d894652e0c9e257f9 +size 3295359 diff --git a/masters/abc-sim/spell_fish/video.mp4 b/masters/abc-sim/spell_fish/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4acc7accc21d422c270cb2cd39985e386e74bf7a --- /dev/null +++ b/masters/abc-sim/spell_fish/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfd4de828addf238c58b8bb27a2fc81fd54419c0bab3beae39d32ce726b68820 +size 10139813 diff --git a/masters/abc-sim/spell_xdof/meta.json b/masters/abc-sim/spell_xdof/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..e684adf14c57890e109d74878ef91f23999642b4 --- /dev/null +++ b/masters/abc-sim/spell_xdof/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_xdof", + "task_dir": "abc-sim-spell-xdof-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-xdof-i00-privileged/abc-sim-spell-xdof-i00-privilege__XdSs2XG", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-xdof-i00-privileged/abc-sim-spell-xdof-i00-privilege__XdSs2XG", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 683, + "seconds": 22.77, + "speed": 4.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (aca36cebe6bea90b), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 2437, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 2437, + "control_steps_replayed": 2437, + "sim_time_s": 82.858, + "frames_every_control_steps": 3.9216, + "final_state_sha": "aca36cebe6bea90b", + "plain_replay_state_sha": "aca36cebe6bea90b", + "replays_agree": true, + "instance_sha": "2566d3e2b18fb048", + "init_sha": "2566d3e2b18fb048", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "cabd9abe2ddeccf55952ad50a73409d1cf0a932381e69d929201ec573a47bcec", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "aca36cebe6bea90b" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5152, + 0.0339, + 0.8792 + ], + "distance_m": 1.7923, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 5.08, + "created": "2026-10-31T05:25:34-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_xdof/poster.png b/masters/abc-sim/spell_xdof/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..8e04934eb9620c891977772d169a3db382c9ce3e --- /dev/null +++ b/masters/abc-sim/spell_xdof/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bbd27cce8db14f0cb4d17f94f4a47983896f66ba2c5a20d6867869fea3dc24f7 +size 3275751 diff --git a/masters/abc-sim/spell_xdof/video.mp4 b/masters/abc-sim/spell_xdof/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b18ee3fe42259aab80aaf77e1dbb09e2d92272c4 --- /dev/null +++ b/masters/abc-sim/spell_xdof/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:430f40d3379757d1c362424882287efa05698b8bcc1b8771319416f8690bc6a4 +size 10858742 diff --git a/masters/abc-sim/spell_yam/meta.json b/masters/abc-sim/spell_yam/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..098ecade744e0a2338e5f6b14a8b786fc8829a9c --- /dev/null +++ b/masters/abc-sim/spell_yam/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "spell_yam", + "task_dir": "abc-sim-spell-yam-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-spell-yam-i00-privileged/abc-sim-spell-yam-i00-privileged__ejiGWEV", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-spell-yam-i00-privileged/abc-sim-spell-yam-i00-privileged__ejiGWEV", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 674, + "seconds": 22.47, + "speed": 4.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (32473fca74b032c6), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 2402, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 2402, + "control_steps_replayed": 2402, + "sim_time_s": 81.668, + "frames_every_control_steps": 3.9216, + "final_state_sha": "32473fca74b032c6", + "plain_replay_state_sha": "32473fca74b032c6", + "replays_agree": true, + "instance_sha": "f9108c924a91c829", + "init_sha": "f9108c924a91c829", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "7453ac4955c64c489482dcf167171a3d512232ee2cd39ac05ff5d99d83f209af", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "32473fca74b032c6" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4926, + 0.0156, + 0.8941 + ], + "distance_m": 1.695, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 7.66, + "created": "2026-10-31T05:38:22-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/spell_yam/poster.png b/masters/abc-sim/spell_yam/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..9ca5addc9bd1f9b8f7b7f4e8074e8053c3265c67 --- /dev/null +++ b/masters/abc-sim/spell_yam/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c3d21a73ffa6b201bc9471e981f456032c6a831efaae45790bf0911c786aaa1 +size 3350378 diff --git a/masters/abc-sim/spell_yam/video.mp4 b/masters/abc-sim/spell_yam/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a195de243478edce4d90c2248a28bc4e6e3a9405 --- /dev/null +++ b/masters/abc-sim/spell_yam/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0d2cd88e2fb6a1032b621790ecf2ee3512dd9a69f991b77e2c0702506e1a7d6 +size 11125996 diff --git a/masters/abc-sim/sweep/meta.json b/masters/abc-sim/sweep/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..c678b4f7a72ce581b0d9835f86e3320fdc100d12 --- /dev/null +++ b/masters/abc-sim/sweep/meta.json @@ -0,0 +1,79 @@ +{ + "suite": "abc-sim", + "task": "sweep", + "task_dir": "abc-sim-sweep-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-sweep-i00-privileged/abc-sim-sweep-i00-privileged__xEUjE5i", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-sweep-i00-privileged/abc-sim-sweep-i00-privileged__xEUjE5i", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 620, + "seconds": 20.67, + "speed": 1.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (2b8e7077f5f0332a), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 548, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 548, + "control_steps_replayed": 548, + "sim_time_s": 18.632, + "frames_every_control_steps": 0.9804, + "final_state_sha": "2b8e7077f5f0332a", + "plain_replay_state_sha": "2b8e7077f5f0332a", + "replays_agree": true, + "instance_sha": "acdf58bdd8888ca2", + "init_sha": "acdf58bdd8888ca2", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "b0f1a5229cf2278e33c8e1a87233ec0df482724cf9e2f9ef64dbcceb3e92d069", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "2b8e7077f5f0332a" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)", + "unused objects the scene parks far below or above the table (they fall freely and touch nothing) not drawn: trash_3, trash_4, trash_5, trash_6, trash_7" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4952, + 0.023, + 0.96 + ], + "distance_m": 1.5399, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 8.58, + "created": "2026-10-31T07:35:04-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/sweep/poster.png b/masters/abc-sim/sweep/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..7b19a2005880ac3c98031707597c7c5bf594704d --- /dev/null +++ b/masters/abc-sim/sweep/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ecc8de759bb98c77709cf9f81daa6ebb9e42c9020bbb4c3a0d17cb89ee4ffe4 +size 3305226 diff --git a/masters/abc-sim/sweep/video.mp4 b/masters/abc-sim/sweep/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9b7b25d9392b7ee7c0c812c9994dd7097962dddc --- /dev/null +++ b/masters/abc-sim/sweep/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ff391ff6c3503accbb01811282d4b2eef020ac1a6fd0950ac8a4820cc35168c +size 10396572 diff --git a/masters/abc-sim/teaser.mp4 b/masters/abc-sim/teaser.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4341711bc9092a878dc9597aaa3da230118a2676 --- /dev/null +++ b/masters/abc-sim/teaser.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd0a43fb19da550869df6db66138c5bfc3c132e8babf8d845989a660948c6f10 +size 9892185 diff --git a/masters/abc-sim/teaser_poster.png b/masters/abc-sim/teaser_poster.png new file mode 100644 index 0000000000000000000000000000000000000000..2b6cd2e23fd8353b21cd1f22b401ec2efec4b5d1 --- /dev/null +++ b/masters/abc-sim/teaser_poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:370911b4da00e4baaeda83f07d69856f12d4ec44864527df2a4176f95c8d9517 +size 3237368 diff --git a/masters/abc-sim/throw_bottles/meta.json b/masters/abc-sim/throw_bottles/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..f5a96849a72d497363b365622db694666ed3ba92 --- /dev/null +++ b/masters/abc-sim/throw_bottles/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "throw_bottles", + "task_dir": "abc-sim-throw-bottles-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-throw-bottles-i00-privileged/abc-sim-throw-bottles-i00-privil__eNR4uF3", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-throw-bottles-i00-privileged/abc-sim-throw-bottles-i00-privil__eNR4uF3", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 621, + "seconds": 20.7, + "speed": 5.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (a4c779ea4cc7deb3), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 2742, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 2742, + "control_steps_replayed": 2742, + "sim_time_s": 93.228, + "frames_every_control_steps": 4.902, + "final_state_sha": "a4c779ea4cc7deb3", + "plain_replay_state_sha": "a4c779ea4cc7deb3", + "replays_agree": true, + "instance_sha": "aadad7bdb8c02a78", + "init_sha": "aadad7bdb8c02a78", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "e49f6ece0315481cd88d447c8e222c4b2295c9fa59b3cc52f924f6c0d8649636", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "a4c779ea4cc7deb3" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.5533, + -0.0803, + 0.8783 + ], + "distance_m": 1.8919, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 4.6, + "created": "2026-10-31T04:29:14-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/throw_bottles/poster.png b/masters/abc-sim/throw_bottles/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..5884707661455e869905fafdcf14cd66ee0c70a6 --- /dev/null +++ b/masters/abc-sim/throw_bottles/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b23233dc5f99d3aeedddf724fcdcc1fa8888acfc2c29e6766e71b2225dbb5387 +size 3152671 diff --git a/masters/abc-sim/throw_bottles/video.mp4 b/masters/abc-sim/throw_bottles/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4977712f6253256e384d2119bf46d3dbfa618daa --- /dev/null +++ b/masters/abc-sim/throw_bottles/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d724baf3b0a6b4979cdb45ef6962b381478258c0671326c495d341230ad46cf7 +size 9242921 diff --git a/masters/abc-sim/turn_mug/meta.json b/masters/abc-sim/turn_mug/meta.json new file mode 100644 index 0000000000000000000000000000000000000000..475554e6ae51d35ea36e28157beff0f9bd6eb04c --- /dev/null +++ b/masters/abc-sim/turn_mug/meta.json @@ -0,0 +1,78 @@ +{ + "suite": "abc-sim", + "task": "turn_mug", + "task_dir": "abc-sim-turn-mug-i00-privileged", + "source": "agent-success:codex-gpt6_luna-xhigh-abc-sim-turn-mug-i00-privileged/abc-sim-turn-mug-i00-privileged__6fGE6be", + "agent": "codex + GPT-6 Luna (xhigh)", + "run": "codex-gpt6_luna-xhigh", + "mode": "privileged", + "trial": "codex-gpt6_luna-xhigh-abc-sim-turn-mug-i00-privileged/abc-sim-turn-mug-i00-privileged__6fGE6be", + "trial_choice": "GPT-6 Luna's counted success; privileged before standard (a privileged trial is the submitted trajectory, replayed after the run)", + "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives, colours and UV textures); seamless studio floor (cyclorama), one large soft key area light, a weak overhead softbox, a rim light, a dim sky dome (Poly Haven studio_small_09 HDRI, CC0); AgX view transform, Medium High Contrast look; fixed 3/4 camera with a slow eased 10-degree orbit; native 2560x1440", + "rendering": "path-tracing", + "media_standard": "v2 (2026-10-16)", + "width": 2560, + "height": 1440, + "fps": 30, + "frames": 713, + "seconds": 23.77, + "speed": 3.0, + "success_check": true, + "success_check_how": "privileged trial: the submitted trajectory (trajectory.npz) replayed from the task's frozen instance with the task's own tooling (rcb_abc.Session, as tests' grade.py replays it): success by the task's rule (upstream's evaluator held 30 control steps), the end-state hash equals both grading replays' hash (c271c38373ea15e1), and a second plain replay in a fresh session (Session.replay) ends in the same hash", + "end_check": { + "success": true, + "rule": "event", + "reason": "success", + "steps": 1916, + "success_run": 30, + "hold_steps": 30 + }, + "replay": { + "control_hz": 29.4118, + "control_dt_s": 0.034, + "n_actions": 1916, + "control_steps_replayed": 1916, + "sim_time_s": 65.144, + "frames_every_control_steps": 2.9412, + "final_state_sha": "c271c38373ea15e1", + "plain_replay_state_sha": "c271c38373ea15e1", + "replays_agree": true, + "instance_sha": "daf231053bf05830", + "init_sha": "daf231053bf05830", + "instance_ok": true, + "state_hashes_every_control_steps": 1000, + "trajectory_sha256": "3d591a35991269a2f9ec73c065ff8c43a76695c427aabd0529a20cbae3d0d773", + "mujoco": "3.6.0", + "recorded_grader_state_sha": "c271c38373ea15e1" + }, + "visual_changes": [ + "studio floor and cove instead of MuJoCo's floor and the booth's three walls", + "the overhead camera gate (bar and camera) not drawn", + "the table top drawn as a solid light-wood block down to the floor, extended under the arms' base mounts (top face unchanged; the scene's table is a thin white slab)" + ], + "camera": { + "mode": "fixed", + "frames_on": "both arms and every object on the table", + "az_deg": 35.0, + "el_deg": 32.0, + "fovy_deg": 34.0, + "fill": 0.8, + "target": [ + 0.4712, + 0.0203, + 0.9209 + ], + "distance_m": 1.4748, + "orbit_deg": 10.0 + }, + "holds": { + "start_s": 0.5, + "end_s": 1.5, + "note": "first / last frame held, no extra simulation" + }, + "image": "efe:abc-sim-0.1.0", + "encode": "libx264 preset slow CRF 16, yuv420p, bt709, faststart", + "poster": "last frame (solved state)", + "seconds_per_frame": 5.02, + "created": "2026-10-31T05:25:39-0500" +} \ No newline at end of file diff --git a/masters/abc-sim/turn_mug/poster.png b/masters/abc-sim/turn_mug/poster.png new file mode 100644 index 0000000000000000000000000000000000000000..11dcd6eac2c1fb89a575f494094c0d276852c332 --- /dev/null +++ b/masters/abc-sim/turn_mug/poster.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e04d89e5672855ea1ad1ce646aafe6b379cd9386fac4e67334d3dc092b9dc85 +size 3247717 diff --git a/masters/abc-sim/turn_mug/video.mp4 b/masters/abc-sim/turn_mug/video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7c01f2b1e8c96c28f056d804cc415a11218b1f91 --- /dev/null +++ b/masters/abc-sim/turn_mug/video.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32a35186be07648706db6988620854a816fbe995b2c068b46078569e6f9bf680 +size 11334079