pns-bind-25m / e8_full /reference /CONFIRMATION_GATE.json
nur-dev's picture
Add E8-FULL reproducibility assets and update author metadata
b9461a3 verified
Raw History Blame Contribute Delete
23.7 kB
{
"all_passed": true,
"checkpoints": {
"grounded_cut_s821": "cf7099132e28e2c8789bcee308de3a6bf7ae8f428d00cd03b6e2c9aff2fb0dda",
"grounded_cut_s822": "d87765232da523d8127bd53895ad485942c1bd1ee12c17205fa41519af1946f7",
"grounded_cut_s823": "ca95bfca612b3e07ff0ea13a5112b9d5add1178bff88fc4f0baf8197ddb26746",
"grounded_future_s821": "82e6ade42e299b16cb3088974429fcac1c28da125a959b2d3a0ed9ac9ab6e571",
"grounded_future_s822": "d6b28b958d38cb0e6ede657816627180658dcd9e119da7ed887c4de470ccc9fc",
"grounded_future_s823": "e99e4518412442a35dd74d64883051086eda56f97d22ae3a33c5a330c50d7ced",
"numeric_future_s821": "43edca0e292e714aeeceab2b21bc1a40f33084fbfb6fd89e3c7c6dc901ec482b",
"numeric_future_s822": "f3497059dda69a72e63e2dbd35b84fe377a8aaa762825e275b41aea9f70d0da3",
"numeric_future_s823": "3de69de35bfb8d3c715adb9d90ac1b25e56ad1458cfeb54b255b76fc62cace45"
},
"effects": {
"821": {
"grounding": {
"delta": 0.9385,
"paired_95_ci": [
0.928,
0.9485
]
},
"ordinary_families": {
"0": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 13814,
"paired_vs_numeric_95_ci": [
0.0,
0.0
]
},
"1": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 7218,
"paired_vs_numeric_95_ci": [
0.43684768548703923,
0.46582393882626855
]
},
"10": {
"accuracy": 0.805360561582642,
"lifetime_bootstrap_95_ci": [
0.797574747836381,
0.8133098737234353
],
"n": 10969,
"paired_vs_numeric_95_ci": [
0.0327356555760956,
0.047053449063499314
]
},
"2": {
"accuracy": 0.7300678354330015,
"lifetime_bootstrap_95_ci": [
0.7244758130953999,
0.7355560693070885
],
"n": 22702,
"paired_vs_numeric_95_ci": [
0.023215525231701117,
0.031436994892015714
]
},
"3": {
"accuracy": 0.6404595225273739,
"lifetime_bootstrap_95_ci": [
0.6249988827314981,
0.6556016962542819
],
"n": 5571,
"paired_vs_numeric_95_ci": [
-0.006816757012024072,
0.00874651947199673
]
},
"4": {
"accuracy": 0.8765338256053861,
"lifetime_bootstrap_95_ci": [
0.8700413835175018,
0.8829439102713796
],
"n": 9209,
"paired_vs_numeric_95_ci": [
0.04324237548388615,
0.05280316922508261
]
},
"5": {
"accuracy": 0.8566287439052704,
"lifetime_bootstrap_95_ci": [
0.8489120893527262,
0.8640732330529396
],
"n": 8614,
"paired_vs_numeric_95_ci": [
0.0407904335359734,
0.05073727637746597
]
},
"6": {
"accuracy": 0.5002173598029271,
"lifetime_bootstrap_95_ci": [
0.49057595560617867,
0.5096733815865663
],
"n": 6901,
"paired_vs_numeric_95_ci": [
-0.009746308037313729,
0.016087671879848057
]
},
"7": {
"accuracy": 0.4966454209996645,
"lifetime_bootstrap_95_ci": [
0.4885080599962963,
0.504528795703886
],
"n": 5962,
"paired_vs_numeric_95_ci": [
-0.0023558387955821608,
0.015258407937981857
]
},
"8": {
"accuracy": 0.37179992122883027,
"lifetime_bootstrap_95_ci": [
0.3529334381352303,
0.3900368673926125
],
"n": 7617,
"paired_vs_numeric_95_ci": [
-0.011837588368586609,
0.022985746338649634
]
},
"9": {
"accuracy": 0.5021,
"lifetime_bootstrap_95_ci": [
0.4956,
0.5080025
],
"n": 10000,
"paired_vs_numeric_95_ci": [
-0.0085,
0.0079
]
}
},
"relation_zero": {
"delta": 0.9195,
"paired_95_ci": [
0.907,
0.9305
]
},
"second_zero": {
"delta": 0.9695,
"paired_95_ci": [
0.962,
0.9765
]
},
"writer_credit": {
"delta": 0.5755,
"paired_95_ci": [
0.5544875,
0.5960125
]
}
},
"822": {
"grounding": {
"delta": 0.935,
"paired_95_ci": [
0.9235,
0.9455
]
},
"ordinary_families": {
"0": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 13814,
"paired_vs_numeric_95_ci": [
0.0,
0.0
]
},
"1": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 7218,
"paired_vs_numeric_95_ci": [
0.4434360984477531,
0.4733765089668616
]
},
"10": {
"accuracy": 0.802260917130094,
"lifetime_bootstrap_95_ci": [
0.7947425799286487,
0.8103368000162623
],
"n": 10969,
"paired_vs_numeric_95_ci": [
0.026219966146918468,
0.03961661673117638
]
},
"2": {
"accuracy": 0.7294071006959739,
"lifetime_bootstrap_95_ci": [
0.723459794390412,
0.7350646793443505
],
"n": 22702,
"paired_vs_numeric_95_ci": [
0.02285196008499784,
0.031895419916613744
]
},
"3": {
"accuracy": 0.6372285047567762,
"lifetime_bootstrap_95_ci": [
0.6222776710863696,
0.6523274043216719
],
"n": 5571,
"paired_vs_numeric_95_ci": [
-0.006134997005036631,
0.008792549770449106
]
},
"4": {
"accuracy": 0.8743620371375828,
"lifetime_bootstrap_95_ci": [
0.8678399987848134,
0.8807406356456956
],
"n": 9209,
"paired_vs_numeric_95_ci": [
0.041753407677607796,
0.05159216572078535
]
},
"5": {
"accuracy": 0.8540747620153238,
"lifetime_bootstrap_95_ci": [
0.846622423894907,
0.861738713735488
],
"n": 8614,
"paired_vs_numeric_95_ci": [
0.03385819983419254,
0.04471857679255813
]
},
"6": {
"accuracy": 0.5010867990146356,
"lifetime_bootstrap_95_ci": [
0.49236318208077307,
0.5101722897012719
],
"n": 6901,
"paired_vs_numeric_95_ci": [
-0.0101243853051779,
0.01732245853909791
]
},
"7": {
"accuracy": 0.5008386447500839,
"lifetime_bootstrap_95_ci": [
0.4934628471098047,
0.5082260210614573
],
"n": 5962,
"paired_vs_numeric_95_ci": [
-0.010734832973158075,
0.003860436883144213
]
},
"8": {
"accuracy": 0.36996192726795324,
"lifetime_bootstrap_95_ci": [
0.35252967222454046,
0.3874115279276518
],
"n": 7617,
"paired_vs_numeric_95_ci": [
-0.02302871790490305,
0.014112435891901467
]
},
"9": {
"accuracy": 0.4981,
"lifetime_bootstrap_95_ci": [
0.492,
0.5039025
],
"n": 10000,
"paired_vs_numeric_95_ci": [
-0.0085025,
0.008302499999999987
]
}
},
"relation_zero": {
"delta": 0.9195,
"paired_95_ci": [
0.907,
0.9305
]
},
"second_zero": {
"delta": 0.9755,
"paired_95_ci": [
0.9685,
0.982
]
},
"writer_credit": {
"delta": 0.5825,
"paired_95_ci": [
0.5609875000000001,
0.605
]
}
},
"823": {
"grounding": {
"delta": 0.94,
"paired_95_ci": [
0.929,
0.95
]
},
"ordinary_families": {
"0": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 13814,
"paired_vs_numeric_95_ci": [
0.0,
0.0005800916188892
]
},
"1": {
"accuracy": 1.0,
"lifetime_bootstrap_95_ci": [
1.0,
1.0
],
"n": 7218,
"paired_vs_numeric_95_ci": [
0.43410743467236745,
0.46507646871810315
]
},
"10": {
"accuracy": 0.8075485459020877,
"lifetime_bootstrap_95_ci": [
0.8001993797602215,
0.8151366206940472
],
"n": 10969,
"paired_vs_numeric_95_ci": [
0.0338572621723962,
0.048068295626372014
]
},
"2": {
"accuracy": 0.7291868557836314,
"lifetime_bootstrap_95_ci": [
0.7234498652453343,
0.734846106044913
],
"n": 22702,
"paired_vs_numeric_95_ci": [
0.02381144370745526,
0.032737013920597205
]
},
"3": {
"accuracy": 0.6476395620175911,
"lifetime_bootstrap_95_ci": [
0.6330665582655827,
0.6619872931717187
],
"n": 5571,
"paired_vs_numeric_95_ci": [
0.0005371338805813189,
0.014674499134543967
]
},
"4": {
"accuracy": 0.8653491149961994,
"lifetime_bootstrap_95_ci": [
0.8587660044757502,
0.8717810670262505
],
"n": 9209,
"paired_vs_numeric_95_ci": [
0.03188806772536158,
0.04021232535857154
]
},
"5": {
"accuracy": 0.8479219874622708,
"lifetime_bootstrap_95_ci": [
0.8402263453372459,
0.8557715296003527
],
"n": 8614,
"paired_vs_numeric_95_ci": [
0.02999628083505561,
0.03989356097506071
]
},
"6": {
"accuracy": 0.5041298362556151,
"lifetime_bootstrap_95_ci": [
0.49642900912403853,
0.5123931004493405
],
"n": 6901,
"paired_vs_numeric_95_ci": [
-0.00403112606002062,
0.001596198660039464
]
},
"7": {
"accuracy": 0.5010063737001006,
"lifetime_bootstrap_95_ci": [
0.49379595380270885,
0.5085485845487169
],
"n": 5962,
"paired_vs_numeric_95_ci": [
-0.004024819721616636,
0.003861895865647299
]
},
"8": {
"accuracy": 0.4018642510174609,
"lifetime_bootstrap_95_ci": [
0.38319438731499506,
0.4214750692655941
],
"n": 7617,
"paired_vs_numeric_95_ci": [
-0.0006545380763896737,
0.030161156756118964
]
},
"9": {
"accuracy": 0.5006,
"lifetime_bootstrap_95_ci": [
0.4962,
0.505
],
"n": 10000,
"paired_vs_numeric_95_ci": [
-0.0047025,
0.003
]
}
},
"relation_zero": {
"delta": 0.9195,
"paired_95_ci": [
0.907,
0.9305
]
},
"second_zero": {
"delta": 0.9755,
"paired_95_ci": [
0.9685,
0.982
]
},
"writer_credit": {
"delta": 0.619,
"paired_95_ci": [
0.598,
0.641
]
}
}
},
"seeds": {
"821": {
"all_passed": true,
"gates": {
"challenge_drop_relation_zero": true,
"challenge_drop_second_zero": true,
"challenge_identity_0": true,
"challenge_identity_128": true,
"challenge_identity_512": true,
"challenge_n": true,
"challenge_onehop_0": true,
"challenge_onehop_128": true,
"challenge_onehop_512": true,
"challenge_oracle": true,
"challenge_permutation_agrees": true,
"challenge_relation_cf_answer": true,
"challenge_relation_cf_identity": true,
"challenge_replay_agrees": true,
"challenge_restore_agrees": true,
"challenge_twohop_0": true,
"challenge_twohop_128": true,
"challenge_twohop_512": true,
"challenge_twohop_ci_0": true,
"challenge_twohop_ci_128": true,
"challenge_twohop_ci_512": true,
"challenge_value_cf_answer": true,
"current_only": true,
"family_10_power": true,
"family_10_preserved": true,
"family_2_power": true,
"family_2_preserved": true,
"family_3_power": true,
"family_3_preserved": true,
"family_4_power": true,
"family_4_preserved": true,
"family_5_power": true,
"family_5_preserved": true,
"family_6_power": true,
"family_6_preserved": true,
"family_7_power": true,
"family_7_preserved": true,
"family_8_power": true,
"family_8_preserved": true,
"family_9_power": true,
"family_9_preserved": true,
"identity_advantage_grounded_cut": true,
"identity_advantage_numeric_future": true,
"identity_retention": true,
"long_replay": true,
"long_restore": true,
"ordinary_n": true,
"pns_current_margin": true,
"pns_drop_relation_zero": true,
"pns_drop_second_zero": true,
"pns_identity": true,
"pns_latest": true,
"pns_long_acc": true,
"pns_long_delay": true,
"pns_long_power": true,
"pns_oracle": true,
"pns_replay": true,
"pns_restore": true,
"pns_twohop": true,
"twohop_retention": true,
"unrelated_changed": true
},
"unmeasured": []
},
"822": {
"all_passed": true,
"gates": {
"challenge_drop_relation_zero": true,
"challenge_drop_second_zero": true,
"challenge_identity_0": true,
"challenge_identity_128": true,
"challenge_identity_512": true,
"challenge_n": true,
"challenge_onehop_0": true,
"challenge_onehop_128": true,
"challenge_onehop_512": true,
"challenge_oracle": true,
"challenge_permutation_agrees": true,
"challenge_relation_cf_answer": true,
"challenge_relation_cf_identity": true,
"challenge_replay_agrees": true,
"challenge_restore_agrees": true,
"challenge_twohop_0": true,
"challenge_twohop_128": true,
"challenge_twohop_512": true,
"challenge_twohop_ci_0": true,
"challenge_twohop_ci_128": true,
"challenge_twohop_ci_512": true,
"challenge_value_cf_answer": true,
"current_only": true,
"family_10_power": true,
"family_10_preserved": true,
"family_2_power": true,
"family_2_preserved": true,
"family_3_power": true,
"family_3_preserved": true,
"family_4_power": true,
"family_4_preserved": true,
"family_5_power": true,
"family_5_preserved": true,
"family_6_power": true,
"family_6_preserved": true,
"family_7_power": true,
"family_7_preserved": true,
"family_8_power": true,
"family_8_preserved": true,
"family_9_power": true,
"family_9_preserved": true,
"identity_advantage_grounded_cut": true,
"identity_advantage_numeric_future": true,
"identity_retention": true,
"long_replay": true,
"long_restore": true,
"ordinary_n": true,
"pns_current_margin": true,
"pns_drop_relation_zero": true,
"pns_drop_second_zero": true,
"pns_identity": true,
"pns_latest": true,
"pns_long_acc": true,
"pns_long_delay": true,
"pns_long_power": true,
"pns_oracle": true,
"pns_replay": true,
"pns_restore": true,
"pns_twohop": true,
"twohop_retention": true,
"unrelated_changed": true
},
"unmeasured": []
},
"823": {
"all_passed": true,
"gates": {
"challenge_drop_relation_zero": true,
"challenge_drop_second_zero": true,
"challenge_identity_0": true,
"challenge_identity_128": true,
"challenge_identity_512": true,
"challenge_n": true,
"challenge_onehop_0": true,
"challenge_onehop_128": true,
"challenge_onehop_512": true,
"challenge_oracle": true,
"challenge_permutation_agrees": true,
"challenge_relation_cf_answer": true,
"challenge_relation_cf_identity": true,
"challenge_replay_agrees": true,
"challenge_restore_agrees": true,
"challenge_twohop_0": true,
"challenge_twohop_128": true,
"challenge_twohop_512": true,
"challenge_twohop_ci_0": true,
"challenge_twohop_ci_128": true,
"challenge_twohop_ci_512": true,
"challenge_value_cf_answer": true,
"current_only": true,
"family_10_power": true,
"family_10_preserved": true,
"family_2_power": true,
"family_2_preserved": true,
"family_3_power": true,
"family_3_preserved": true,
"family_4_power": true,
"family_4_preserved": true,
"family_5_power": true,
"family_5_preserved": true,
"family_6_power": true,
"family_6_preserved": true,
"family_7_power": true,
"family_7_preserved": true,
"family_8_power": true,
"family_8_preserved": true,
"family_9_power": true,
"family_9_preserved": true,
"identity_advantage_grounded_cut": true,
"identity_advantage_numeric_future": true,
"identity_retention": true,
"long_replay": true,
"long_restore": true,
"ordinary_n": true,
"pns_current_margin": true,
"pns_drop_relation_zero": true,
"pns_drop_second_zero": true,
"pns_identity": true,
"pns_latest": true,
"pns_long_acc": true,
"pns_long_delay": true,
"pns_long_power": true,
"pns_oracle": true,
"pns_replay": true,
"pns_restore": true,
"pns_twohop": true,
"twohop_retention": true,
"unrelated_changed": true
},
"unmeasured": []
}
},
"sources": {
"campaigns/e8_full/PROTOCOL.md": "d963c30b0012f713d1aa37d25384af26178008c8a9e9b99b66539c750c595ec5",
"campaigns/e8_full/benchmark.py": "2875fa4f617210201fbbb18ed63bd37e6084a4c7e43b71627cd61356097fe172",
"campaigns/e8_full/common.py": "5ef5ca21fe57da06cbf020c20d32c46b75ef52463abeb6e8c7375d79591c9843",
"campaigns/e8_full/data.py": "276537eca7aa7cbd45cc18fc7d306289a01344af18a154c2168ea70393a70460",
"campaigns/e8_full/evaluate.py": "91cd979a26384ed63be3acc36b401860bb522cefa41af6874b3b6e8daff41bf3",
"campaigns/e8_full/gates.py": "df17f274184cb91b4c9023949fa896e7df662fea6d11208d2cd48aeaad249f3b",
"campaigns/e8_full/model.py": "d06963e144e4536b599727891af9776ee12c13d4c4a523753655e04051c51b8c",
"campaigns/e8_full/preflight.py": "41bd0f5ed3b6439ea2d12f29364c266c1f32506a46cc18f396d8d00f1d114c97",
"campaigns/e8_full/run_campaign.py": "170ba0d116c8186f9223d5b0d17a0ed9f7ae7949ca7618e700a30cb472c69ae9",
"campaigns/e8_full/train.py": "3fabd69993fe4523d4925556d299b8620387e8c0d5369c85a04a9c89c331e3e6",
"campaigns/e8_identity_future/PREREGISTRATION_FULL.md": "7a5ac518262079cc3175f59ef12c01d25d6011e1af28efad26b3fb0f74301bf6",
"campaigns/e8_identity_future/core.py": "33905ffd455385a657c066648349e0595aa165daecd61380b7e8c52b47968425",
"pns/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/common.py": "c86ea65aae22266b52e2349d9589115b75b6881a4a88eabb18fd122e319b4aa4",
"pns/data/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/data/binding.py": "b5721cbf0f14bad2522590f86214b6b541b20fff52e91b63079737b1dc3fd56e",
"pns/data/pack.py": "6b26767b2173a825eb023768b105e2f83d3cdc8b97603eb190f4bb344ce8e582",
"pns/data/tokenize.py": "bd550ac2a73ce7f5180c4bdcc67e782af86c657e8dec5e1687ab479840ab934a",
"pns/data/view.py": "ee90a4c3e966efcc026ca6180e6e3291b4eba21a9df32c37dbd9ab9c7f36e88a",
"pns/derange.py": "0791dd8a7478746271fffa85ef52f7ccf7f4f50492dc79bfcec20464a909daa9",
"pns/eval/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/model/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/model/bind.py": "d18c84785e4456e41bb9962eaf91f97063fad87e729bdc7a284439ca66c80f8c",
"pns/model/compose.py": "f825342fb28cda25028cca6b9d3ff484f6a57dc4b25de3b8c3ebaf0a7ab4370c",
"pns/model/frozenhead.py": "71028c51cc4a709a50e51aa7703487c527af78550ffc7ca807ea5e2271685aeb",
"pns/model/modules.py": "15938c7dcc34828e12358b90927b91aaa09fb666099d5d4c1e6b397afd8e483c",
"pns/model/pnsr.py": "15700dae63b7bccb6dd32961f3f9106feb724d98ef74efdf48e8aeb611fbc7ed",
"pns/model/relrep.py": "2f4be9f90c0780619f5fcf7407792f32831eb9688720877c96ff3c3280540afb",
"pns/model/relwrite.py": "f040b482b96e1fb7280cfac25db65fad8cedee4d524cba007a0e8ae44bbcfcda",
"pns/model/rmt.py": "d64cb291a26a757c372c8cbb51a344823cbc2c8b74947a2e4e57a99e303c3263",
"pns/model/tx.py": "9c88fbab991b7a3748a662403b119037badf4f7f547c5ec7f6639523f0bcf64c",
"pns/train/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/train/loader.py": "df1fa3f4590bbcdd17f94984b666a6e4ceceaee9e9f92e9c588a0a9b52cbedfd",
"pns/verdict.py": "d43470dc3c2a0f365775866bd2d031b6b52612cf77c6a97d40ea9bd1651f6e74",
"pns/world/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
"pns/world/questions.py": "1062709b7b777afd2b5a4a3e45793654737f785a58251ee64cd14b894ed7234d",
"pns/world/reducer.py": "216ce037b62c4543dd4d6536b2039ed005cffcf2828efa507c800545932c41c6",
"pns/world/schema.py": "45b06d3d2b3e44e4cf9c68c92213b645f0d2f94208bf7e4dd09b2ba95745098e",
"pns/world/sim.py": "19efe037de29b4e835f1f2281489daf20959b554a86d6fa660e955c1dfdb2cae",
"pns/world/verbalize.py": "5fe1c98bac552dbfdcb237dc128c2f5ee94e60eca4dd81a2366264b95ef3ff93",
"tokenizer/tokenizer.json": "8bbb4fd7db2928b323b7cb7ba595e88130f49a717fb7a79b3bd33b4b5e248fcf"
},
"split": "confirmation"
}