diff --git a/language_breakthrough_v1/b34_reasoning_curriculum/b34-reasoning-rung2-20260809T085000Z-c83a0d4e.measurement.signed.json b/language_breakthrough_v1/b34_reasoning_curriculum/b34-reasoning-rung2-20260809T085000Z-c83a0d4e.measurement.signed.json new file mode 100644 index 0000000000000000000000000000000000000000..d523035726c4797bb74cf9254de2b993e5325e94 --- /dev/null +++ b/language_breakthrough_v1/b34_reasoning_curriculum/b34-reasoning-rung2-20260809T085000Z-c83a0d4e.measurement.signed.json @@ -0,0 +1,2926 @@ +{ + "payload": { + "abi": "TRIDENT-B34-REASONING-CURRICULUM-A100X4-GATE-v1-MEASUREMENT", + "audits": { + "corrected_scorer_used": true, + "data_provenance": { + "A0_note": "source identity/license/split metadata remains outside the neural path", + "all_licenses_recorded": true, + "all_revisions_immutable_40hex": true, + "comparator_output_source_present": false, + "passed": true, + "proof_annotations_in_neural_input": false, + "section_7_5_dataset_repositories": [ + "EleutherAI/hendrycks_math", + "allenai/ai2_arc", + "framolfese/2WikiMultihopQA", + "google-research-datasets/mbpp", + "hotpotqa/hotpot_qa", + "openai/gsm8k", + "openai/openai_humaneval" + ], + "training_repo_section_7_5_repo_overlap": 0, + "training_sources": [ + { + "ancestry": "human-authored Wikimedia dump; deterministic markup cleaning", + "config": "20231101.en", + "license": "CC-BY-SA-3.0 AND GFDL; attribution manifest retained", + "name": "wikipedia_en", + "repo": "wikimedia/wikipedia", + "revision": "b04c8d1ceb2f5cd4588862100d08de323dccfbaa", + "split": "train", + "weight": 0.15 + }, + { + "ancestry": "human-authored books; deterministic frame stripping", + "config": "default", + "license": "Project Gutenberg license; item public-domain status retained", + "name": "gutenberg_en", + "repo": "manu/project_gutenberg", + "revision": "164853d214065df26a630ee1ab91a0c39e461caf", + "split": "en", + "weight": 0.55 + }, + { + "ancestry": "human-authored GitHub code; extension selection, long-line filter, and exact whitespace-insensitive dedup only", + "config": null, + "license": "per-row SPDX allowlist: MIT, Apache-2.0, BSD-2-Clause, BSD-3-Clause, ISC, CC0-1.0, Unlicense", + "name": "codeparrot_permissive", + "repo": "codeparrot/github-code", + "revision": "b5661e6b17396364b2bcf8e68977b0d28e1ebd19", + "split": "train", + "weight": 0.2 + }, + { + "ancestry": "human-authored mathematical/formal code selected by hand-crafted language-specific heuristics", + "config": "default", + "license": "underlying-repository SPDX allowlist: MIT, Apache-2.0, BSD-2-Clause, BSD-3-Clause, ISC, CC0-1.0, Unlicense", + "name": "algebraic_stack_permissive", + "repo": "typeof/algebraic-stack", + "revision": "260231ddd1c6382ad1454112b1e756d1ec33ed3e", + "split": "train", + "weight": 0.1 + }, + { + "files": [ + "data/train-00000-of-00002-6176cf3d78f48858.parquet" + ], + "license": "Apache-2.0", + "name": "proofwriter_reasoning_curriculum", + "repo": "tasksource/proofwriter", + "revision": "ee4d22d49b8486786cbc01c0b6f6f6266e56f3bd", + "split": "train" + } + ] + }, + "future_byte_causality": { + "future_max_abs_delta": 30.197410583496094, + "passed": true, + "prefix_and_self_max_abs_delta": 0.0, + "tolerance": 0.0 + }, + "new_deployed_parameters": 0, + "panel": { + "certification_or_comparator_sets_loaded": [], + "report": { + "blocks_per_source": 12, + "opened_after_both_arms_complete": true, + "sha256": "0e1d805e4315c5bd5958ff42e12e40c016ba60e804dd837f7ea3fbd3908edd3a", + "shape": [ + 48, + 1024 + ] + }, + "selection": { + "blocks_per_source": 12, + "selection_only": true, + "sha256": "d1474dda9b932ee0bf721c598e4e301681302bf529a51d4a7001fa5ae7fdaf84", + "shape": [ + 48, + 1024 + ] + }, + "selection_report_block_overlap": 0, + "training_panel_exact_overlap": 0 + }, + "proof_or_verifier_annotations_in_input_record": false, + "section_7_5_or_Kimi_training_leak": false, + "transport_direction_constructed": false + }, + "capability": { + "D8_correct": 0, + "D8_correct_by_task": { + "2wiki": 0, + "arc_challenge": 0, + "gsm8k": 0, + "hotpotqa": 0, + "humaneval": 0, + "math": 0, + "mbpp": 0 + }, + "D8_micro_accuracy": 0.0, + "D8_total": 56, + "bars": { + "all_three_reasoning_task_slopes_nonnegative": { + "observed": { + "arc_challenge": -0.0125, + "gsm8k": 0.0, + "math": 0.0 + }, + "passed": false + }, + "qualification_micro_accuracy_minimum": { + "observed": 0.0, + "observed_correct": 0, + "passed": false, + "required_minimum": 0.07142857142857142 + }, + "reasoning_D8_correct_minimum": { + "observed": 0, + "passed": false, + "required_minimum": 3 + }, + "reasoning_D8_minus_D1_minimum": { + "observed": -0.08333333333333333, + "observed_correct_difference": -2, + "passed": false, + "required_minimum": 0.08333333333333333 + }, + "retrieve_reason_D8_minus_no_retrieval_minimum": { + "observed": 0.0, + "observed_correct_difference": 0, + "passed": false, + "required_minimum": 0.08333333333333333 + } + }, + "capability_used_for_gradient_checkpoint_selection_or_extension": false, + "depth_grid": [ + 0, + 1, + 2, + 4, + 8 + ], + "failed_bars": [ + "qualification_micro_accuracy_minimum", + "reasoning_D8_correct_minimum", + "reasoning_D8_minus_D1_minimum", + "retrieve_reason_D8_minus_no_retrieval_minimum", + "all_three_reasoning_task_slopes_nonnegative" + ], + "manifest": { + "files": [ + { + "bytes": 29505064, + "path": "data/validation-00000-of-00001.parquet", + "repo": "framolfese/2WikiMultihopQA", + "revision": "fe713bfbd1afbca1a65246741a75890405d56a3a", + "rows": 12576, + "sha256": "408e2dbb28edc6c8b9ca3ba0c94d4fc7bf17ffb923766593a3a7f546ab4cba59", + "task": "2wiki" + }, + { + "bytes": 203808, + "path": "ARC-Challenge/test-00000-of-00001.parquet", + "repo": "allenai/ai2_arc", + "revision": "210d026faf9955653af8916fad021475a3f00453", + "rows": 1172, + "sha256": "62f03257e737aed263f55c6abf87c7bb0028a44a6bdd2a26eb1279eb42c1d1e9", + "task": "arc_challenge" + }, + { + "bytes": 419088, + "path": "main/test-00000-of-00001.parquet", + "repo": "openai/gsm8k", + "revision": "740312add88f781978c0658806c59bc2815b9866", + "rows": 1319, + "sha256": "ee7b8da9e381df27b9e3f7758a159ab2bdaa4dbaa910546cbbc47e0cb44e4f59", + "task": "gsm8k" + }, + { + "bytes": 27452575, + "path": "distractor/validation-00000-of-00001.parquet", + "repo": "hotpotqa/hotpot_qa", + "revision": "1908d6afbbead072334abe2965f91bd2709910ab", + "rows": 7405, + "sha256": "c20b638ca82b21d04fe12e14ff417ad05153d4d215a65de54497fca4e972f7c6", + "task": "hotpotqa" + }, + { + "bytes": 83920, + "path": "openai_humaneval/test-00000-of-00001.parquet", + "repo": "openai/openai_humaneval", + "revision": "7dce6050a7d6d172f3cc5c32aa97f52fa1a2e544", + "rows": 164, + "sha256": "2f2871a15fbc95b6c683043359f4ed8e144c5a1c4f24f25f66bc51f598dfcfb6", + "task": "humaneval" + }, + { + "bytes": 353114, + "path": "algebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 1187, + "sha256": "0e04a850dac9cfc758eac0c80afd2d0cc65e29ea72f916c8ac1045b12af937f5", + "task": "math" + }, + { + "bytes": 175032, + "path": "counting_and_probability/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 474, + "sha256": "c324818949c75a6227c0328a8499d58023cbbf6757cfc1736f5ab24cc8d87966", + "task": "math" + }, + { + "bytes": 263835, + "path": "geometry/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 479, + "sha256": "98720f3a8134ad9784fd8951195245264495df03ffc5c9485a9a54ff5ae2431a", + "task": "math" + }, + { + "bytes": 394797, + "path": "intermediate_algebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 903, + "sha256": "e859f72486c951c2017da880541ca729940eaa986b0364eb23e238983df9bfeb", + "task": "math" + }, + { + "bytes": 182003, + "path": "number_theory/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 540, + "sha256": "ba156c9c9fa2040836ba12135fd964e6e67d6bccd8c469ec4a3ad941f8f6fbe9", + "task": "math" + }, + { + "bytes": 267729, + "path": "prealgebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 871, + "sha256": "301e5d8ade5c7b5038032532e498bdc9dca799e7f1061a63f2e7d980803240cc", + "task": "math" + }, + { + "bytes": 242241, + "path": "precalculus/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 546, + "sha256": "da37474ba3761ae5727af3844a7ba963a735bd6a14c65e6137593669a4fe3b9c", + "task": "math" + }, + { + "bytes": 60864, + "path": "sanitized/test-00000-of-00001.parquet", + "repo": "google-research-datasets/mbpp", + "revision": "4bb6404fdc6cacfda99d4ac4205087b89d32030c", + "rows": 257, + "sha256": "e9e9efa2c0d59ef5e55537a9d126b8f875d5ac010a8d75628d76824884e15850", + "task": "mbpp" + } + ], + "final_test_split_eval_only": true, + "items": 56, + "items_per_task": 8, + "manifest": { + "path": "results/capability_yardstick_v1/dataset_manifest.json", + "sha256": "d3f2da346429a60af19ec470026888a3d4770c337dc70e534eebac4822e35190" + }, + "opened_after_training_graph_deleted": true, + "preregistration": { + "path": "results/capability_yardstick_v1/preregistration.signed.json", + "sha256": "e017026c98bc64e5bb05928a3c7dc91dbffa6e6cbd78a3c0528119cf0a4fe970" + } + }, + "no_retrieval_D8_correct": 0, + "reasoning_D1_correct": 2, + "reasoning_D8_correct": 0, + "reasoning_accuracy_by_task_and_depth": { + "arc_challenge": { + "0": 0.0, + "1": 0.25, + "2": 0.25, + "4": 0.25, + "8": 0.0 + }, + "gsm8k": { + "0": 0.0, + "1": 0.0, + "2": 0.0, + "4": 0.0, + "8": 0.0 + }, + "math": { + "0": 0.0, + "1": 0.0, + "2": 0.0, + "4": 0.0, + "8": 0.0 + } + }, + "reasoning_slopes": { + "arc_challenge": -0.0125, + "gsm8k": 0.0, + "math": 0.0 + }, + "records": [ + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "b7921a279c82c34a70df709cad46f149e3e187c6a2f99f4b1944f8e0d0d3158e", + "output_sha256": "de3c83a29a77e993f68e1b4a4852a787d0e94137b22d21b0dcfb523f0e9f192b", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "b7921a279c82c34a70df709cad46f149e3e187c6a2f99f4b1944f8e0d0d3158e", + "output_sha256": "de3c83a29a77e993f68e1b4a4852a787d0e94137b22d21b0dcfb523f0e9f192b", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "b7921a279c82c34a70df709cad46f149e3e187c6a2f99f4b1944f8e0d0d3158e", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "b7921a279c82c34a70df709cad46f149e3e187c6a2f99f4b1944f8e0d0d3158e", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "b7921a279c82c34a70df709cad46f149e3e187c6a2f99f4b1944f8e0d0d3158e", + "output_sha256": "23851d3e4262c494b5e26ed64ff4afb9f780fbc8d122134acefbea750a41f71a", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "e71aa0942e9744b32818944d5eec95a709560a5dcb1a8c01cb6ea7ada11a6e38", + "output_sha256": "3b5a6fd40553adfc065abd9cb8f96dd9a9c6d931ec73d43a8866fc4c6479a9d0", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "e71aa0942e9744b32818944d5eec95a709560a5dcb1a8c01cb6ea7ada11a6e38", + "output_sha256": "8e5374c5ef4d988231bb06e3da2b8b7d16fbd382f2c9f6ad88d857c956b7a6dc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "e71aa0942e9744b32818944d5eec95a709560a5dcb1a8c01cb6ea7ada11a6e38", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "e71aa0942e9744b32818944d5eec95a709560a5dcb1a8c01cb6ea7ada11a6e38", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "e71aa0942e9744b32818944d5eec95a709560a5dcb1a8c01cb6ea7ada11a6e38", + "output_sha256": "9e1398a38c5050a8e586fe82deaf8617ab45c162eb9bc34fc120d22ad1e132dd", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "bfe331c856179fba82c69dfb55c258eeb8cc6a06c7050d62b440aadec553cc9c", + "output_sha256": "1b547e8169fc42fcad85adc1b83c410dd2efac1b1f63b39fdf677234b9b662be", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "bfe331c856179fba82c69dfb55c258eeb8cc6a06c7050d62b440aadec553cc9c", + "output_sha256": "806e511d526b13d717def8d83100b4665f673914a06be9e25463903ba2f91cb3", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "bfe331c856179fba82c69dfb55c258eeb8cc6a06c7050d62b440aadec553cc9c", + "output_sha256": "806e511d526b13d717def8d83100b4665f673914a06be9e25463903ba2f91cb3", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "bfe331c856179fba82c69dfb55c258eeb8cc6a06c7050d62b440aadec553cc9c", + "output_sha256": "806e511d526b13d717def8d83100b4665f673914a06be9e25463903ba2f91cb3", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "bfe331c856179fba82c69dfb55c258eeb8cc6a06c7050d62b440aadec553cc9c", + "output_sha256": "201998236020f8d7aa170abd9cc8deec79ad61e7ef62f96dfa600b7dd978d314", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "49cc3ad0432c46791215087ec0806bb4ef211e7ac03d14f646837bba59598fa9", + "output_sha256": "f2a229d721b4ee8fcde6c7146ba0c83fb2fc8cc6aedf54523a619972318d2d70", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "49cc3ad0432c46791215087ec0806bb4ef211e7ac03d14f646837bba59598fa9", + "output_sha256": "f2a229d721b4ee8fcde6c7146ba0c83fb2fc8cc6aedf54523a619972318d2d70", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "49cc3ad0432c46791215087ec0806bb4ef211e7ac03d14f646837bba59598fa9", + "output_sha256": "2263170ed2841f6e0a4d2b32f8dea9941ac2466561fb87e6590c180922ecaf3f", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "49cc3ad0432c46791215087ec0806bb4ef211e7ac03d14f646837bba59598fa9", + "output_sha256": "7ab3fb685466778d19aeb18e92522bae2b03f454bc185fc1240e9c1ca14c7b35", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "49cc3ad0432c46791215087ec0806bb4ef211e7ac03d14f646837bba59598fa9", + "output_sha256": "2a0fbc8200df9f20af23054b59046bacc0ba69991ba087ae155b59357eb545e9", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "e41d1c9f7bee0c8f48c34e8b51d667bc00c7111fb2f1a19e12d5a27a3c396495", + "output_sha256": "11f45cac83602f6fb7bc9b174f0aa87bd49b5296784a2654af5b34f5f6fb6590", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "e41d1c9f7bee0c8f48c34e8b51d667bc00c7111fb2f1a19e12d5a27a3c396495", + "output_sha256": "11f45cac83602f6fb7bc9b174f0aa87bd49b5296784a2654af5b34f5f6fb6590", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "e41d1c9f7bee0c8f48c34e8b51d667bc00c7111fb2f1a19e12d5a27a3c396495", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "e41d1c9f7bee0c8f48c34e8b51d667bc00c7111fb2f1a19e12d5a27a3c396495", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "e41d1c9f7bee0c8f48c34e8b51d667bc00c7111fb2f1a19e12d5a27a3c396495", + "output_sha256": "23851d3e4262c494b5e26ed64ff4afb9f780fbc8d122134acefbea750a41f71a", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "35e2c4f57f2bfaa445ae7cfa6ff5be1935678172a6785216d40c19bd190137f7", + "output_sha256": "bdb32a275652e95b34c34d3d468934abcfe843ab7947bffbc62c487d38df1b76", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "35e2c4f57f2bfaa445ae7cfa6ff5be1935678172a6785216d40c19bd190137f7", + "output_sha256": "bdb32a275652e95b34c34d3d468934abcfe843ab7947bffbc62c487d38df1b76", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "35e2c4f57f2bfaa445ae7cfa6ff5be1935678172a6785216d40c19bd190137f7", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "35e2c4f57f2bfaa445ae7cfa6ff5be1935678172a6785216d40c19bd190137f7", + "output_sha256": "3235fa3edd62fdf853269fac65d4beecf0c2ce4466aaeb148b68b498f56789fc", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "35e2c4f57f2bfaa445ae7cfa6ff5be1935678172a6785216d40c19bd190137f7", + "output_sha256": "82d483c25e474bd1ea68e73289cf3f01631d440d0d804330e0c313ad7ff5a030", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "28d8750962367da7663d71b22b3cfd0f229e2b9dc1e26a305d75d06d0899eb33", + "output_sha256": "7283d637e2df5621e588e84bbb1ebd74d84b72e6de4064c9b0eed7b96dc3bfd4", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "28d8750962367da7663d71b22b3cfd0f229e2b9dc1e26a305d75d06d0899eb33", + "output_sha256": "2bbaa26c45ae067eb6ed1024a5cbf3bf40ae18b1f6dce22bb33ac53124a10ced", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "28d8750962367da7663d71b22b3cfd0f229e2b9dc1e26a305d75d06d0899eb33", + "output_sha256": "2bbaa26c45ae067eb6ed1024a5cbf3bf40ae18b1f6dce22bb33ac53124a10ced", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "28d8750962367da7663d71b22b3cfd0f229e2b9dc1e26a305d75d06d0899eb33", + "output_sha256": "2bbaa26c45ae067eb6ed1024a5cbf3bf40ae18b1f6dce22bb33ac53124a10ced", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "28d8750962367da7663d71b22b3cfd0f229e2b9dc1e26a305d75d06d0899eb33", + "output_sha256": "55eaf245c1dc6b4ab720381866946bb1af293241dcd174a7655eb55541d5432d", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "2ef204d41035a4816606ca241005fab350f31caa40d87102cf545b5cf00ab6b7", + "output_sha256": "05e70371504c1bf6c5749dbb15e621ae529a72b1567814b510a1e61cba27e5e6", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "2ef204d41035a4816606ca241005fab350f31caa40d87102cf545b5cf00ab6b7", + "output_sha256": "77a54ada158287e5cea474ed95b68f751d71fab80f6000dab98664c755bb5386", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "2ef204d41035a4816606ca241005fab350f31caa40d87102cf545b5cf00ab6b7", + "output_sha256": "77a54ada158287e5cea474ed95b68f751d71fab80f6000dab98664c755bb5386", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "2ef204d41035a4816606ca241005fab350f31caa40d87102cf545b5cf00ab6b7", + "output_sha256": "c9dc908c4b49330b1b50bbd3178fc8508b46c1f44cb411c15d1b4e86d8d65d94", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "2ef204d41035a4816606ca241005fab350f31caa40d87102cf545b5cf00ab6b7", + "output_sha256": "bc3274160dcc48070c0ad09d663f4df89badb0fbe592e309c89e07b750e5ad2e", + "pass": false, + "retrieval": true, + "task": "gsm8k" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "74c0820d0fa8b8518e26989f3661cab10cf923d6dea3fc75b40e1f10282fde3e", + "output_sha256": "361e3c32d86cf9cc54baab258185bc804d364b45e74eaf8dbf5a8052a83d16db", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "74c0820d0fa8b8518e26989f3661cab10cf923d6dea3fc75b40e1f10282fde3e", + "output_sha256": "fc0e27f9749d3a5738188e7953e826aa2b2549e39711d442cdadf6295c0328c8", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "74c0820d0fa8b8518e26989f3661cab10cf923d6dea3fc75b40e1f10282fde3e", + "output_sha256": "fc0e27f9749d3a5738188e7953e826aa2b2549e39711d442cdadf6295c0328c8", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "74c0820d0fa8b8518e26989f3661cab10cf923d6dea3fc75b40e1f10282fde3e", + "output_sha256": "fc0e27f9749d3a5738188e7953e826aa2b2549e39711d442cdadf6295c0328c8", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "74c0820d0fa8b8518e26989f3661cab10cf923d6dea3fc75b40e1f10282fde3e", + "output_sha256": "a3f86328963a50db1948a26144abbe4e1e8d06f67117c8ca8aeb77ccc7f8f5c5", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "d88970a32acce4c9d3440f8a8c7551aea87850d7e07be332c2c1d3c79b491a45", + "output_sha256": "faf6099791377bcaf41469cb6d6a9831b4a138647d4c688093eefbab7e790ffc", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "d88970a32acce4c9d3440f8a8c7551aea87850d7e07be332c2c1d3c79b491a45", + "output_sha256": "faf6099791377bcaf41469cb6d6a9831b4a138647d4c688093eefbab7e790ffc", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "d88970a32acce4c9d3440f8a8c7551aea87850d7e07be332c2c1d3c79b491a45", + "output_sha256": "faf6099791377bcaf41469cb6d6a9831b4a138647d4c688093eefbab7e790ffc", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "d88970a32acce4c9d3440f8a8c7551aea87850d7e07be332c2c1d3c79b491a45", + "output_sha256": "a81da10355a83e216a0558fc9def9fb7910814ca3e35fcb486d6fdbbc2ec20ba", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "d88970a32acce4c9d3440f8a8c7551aea87850d7e07be332c2c1d3c79b491a45", + "output_sha256": "c49613f3a24dd8719ecec47eaeb8649cf713ac8145af38a8c297620cc28eb358", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "97fa6a80e36316a7419cf77601a1ada255d0fe6ec93d7371780c53b40d1729a1", + "output_sha256": "77e389e51079dfaf93d67599b810d0cc9de09c7c4fb05bdcfcdcc8effcb491b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "97fa6a80e36316a7419cf77601a1ada255d0fe6ec93d7371780c53b40d1729a1", + "output_sha256": "77e389e51079dfaf93d67599b810d0cc9de09c7c4fb05bdcfcdcc8effcb491b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "97fa6a80e36316a7419cf77601a1ada255d0fe6ec93d7371780c53b40d1729a1", + "output_sha256": "77e389e51079dfaf93d67599b810d0cc9de09c7c4fb05bdcfcdcc8effcb491b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "97fa6a80e36316a7419cf77601a1ada255d0fe6ec93d7371780c53b40d1729a1", + "output_sha256": "e4eba42209b17ad90fbc75c91987ae3b05574d025b8aefc2790aa5548bc3211e", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "97fa6a80e36316a7419cf77601a1ada255d0fe6ec93d7371780c53b40d1729a1", + "output_sha256": "c49613f3a24dd8719ecec47eaeb8649cf713ac8145af38a8c297620cc28eb358", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "98380a0072a4e10feb9d71a28e8a8eaa9ee02a49cdc4ab5096eeeb6eabb96243", + "output_sha256": "f181eb2ac37a9b920649e7dfbf0acf51f9c3f0169c82f34dc42fe2c8d5ede393", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "98380a0072a4e10feb9d71a28e8a8eaa9ee02a49cdc4ab5096eeeb6eabb96243", + "output_sha256": "f181eb2ac37a9b920649e7dfbf0acf51f9c3f0169c82f34dc42fe2c8d5ede393", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "98380a0072a4e10feb9d71a28e8a8eaa9ee02a49cdc4ab5096eeeb6eabb96243", + "output_sha256": "7aebab0d3039f7396c99425dd82d532ed1e6d778711639b43a68a96b75d2f46c", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "98380a0072a4e10feb9d71a28e8a8eaa9ee02a49cdc4ab5096eeeb6eabb96243", + "output_sha256": "ed56f36efbcf150f2b6893108f1286065387961c0c6e8724ec764283350bf8ac", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "98380a0072a4e10feb9d71a28e8a8eaa9ee02a49cdc4ab5096eeeb6eabb96243", + "output_sha256": "4841f4e018c248c96c0ba9fc1317965a21f898de88dc74525a8addfea2e65547", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "f837adb941c8ba2fce457acb52080fc49a3b288e91b0eb22ad8f53deb9e2656a", + "output_sha256": "32de4cce5617b3d8cdb4a847ea122decc6aa2427db694840c88b6c99efc90efe", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "f837adb941c8ba2fce457acb52080fc49a3b288e91b0eb22ad8f53deb9e2656a", + "output_sha256": "c426d6b33e9d49b11b1d327d679da900c6e9c08d9a72815f9050b004e3490a61", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "f837adb941c8ba2fce457acb52080fc49a3b288e91b0eb22ad8f53deb9e2656a", + "output_sha256": "c04f93ce0cce4a966d39c9049408bed064dd2ca24f8119a16c499ffb89f838cf", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "f837adb941c8ba2fce457acb52080fc49a3b288e91b0eb22ad8f53deb9e2656a", + "output_sha256": "165f62f9fa2779678e1432e7699a6092a83d1e0f5c5aeb393f387df504d5c042", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "f837adb941c8ba2fce457acb52080fc49a3b288e91b0eb22ad8f53deb9e2656a", + "output_sha256": "e9e6b5acc3898ccd35538470ee3df19b5d365095d7fd4f3c500e042094558c12", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "218dc3cb6260905ce9c3be838c811548744466824f11ee54daac3d7ffa3217cd", + "output_sha256": "f6a0e97cfabc7f5f2710cc159610f249c47c4ae3ddd4c09b1ec34a72d9bb66a7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "218dc3cb6260905ce9c3be838c811548744466824f11ee54daac3d7ffa3217cd", + "output_sha256": "f6a0e97cfabc7f5f2710cc159610f249c47c4ae3ddd4c09b1ec34a72d9bb66a7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "218dc3cb6260905ce9c3be838c811548744466824f11ee54daac3d7ffa3217cd", + "output_sha256": "f6a0e97cfabc7f5f2710cc159610f249c47c4ae3ddd4c09b1ec34a72d9bb66a7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "218dc3cb6260905ce9c3be838c811548744466824f11ee54daac3d7ffa3217cd", + "output_sha256": "d4146a0e9157255090703cc502e29a85517aee7a5027f4f1c5028685345d8f9f", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "218dc3cb6260905ce9c3be838c811548744466824f11ee54daac3d7ffa3217cd", + "output_sha256": "2f4392615e5c916337c7c4829797681d149317c5161440541770fd344651605d", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "d8268bdc23e78f82ae5fdc6dfd6fcdfe6fb6c4576047df8944fdc14bfaefa55b", + "output_sha256": "3e06cb78ba207b23d8b349bc97c8c92dd897b468104f8f9b601e6eaf5039bf01", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "d8268bdc23e78f82ae5fdc6dfd6fcdfe6fb6c4576047df8944fdc14bfaefa55b", + "output_sha256": "3e06cb78ba207b23d8b349bc97c8c92dd897b468104f8f9b601e6eaf5039bf01", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "d8268bdc23e78f82ae5fdc6dfd6fcdfe6fb6c4576047df8944fdc14bfaefa55b", + "output_sha256": "2755f73193f334c9a045fa391b39be53d451a275f56edacc68e43ce19a99eb07", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "d8268bdc23e78f82ae5fdc6dfd6fcdfe6fb6c4576047df8944fdc14bfaefa55b", + "output_sha256": "75a7229ef3d6a7dd31d3bc0d6a4215fd2ceb691b00a2c0cc908fee9abf9441ee", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "d8268bdc23e78f82ae5fdc6dfd6fcdfe6fb6c4576047df8944fdc14bfaefa55b", + "output_sha256": "c49613f3a24dd8719ecec47eaeb8649cf713ac8145af38a8c297620cc28eb358", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "9dbf1216e37e187c3a89f32c1d0aa3e6d00fb350094818fe192d4a6025311685", + "output_sha256": "c94356c4046e5fe8d00a3dd90111a34bd42b8d2a4508ca606b6448b4ff8735b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "9dbf1216e37e187c3a89f32c1d0aa3e6d00fb350094818fe192d4a6025311685", + "output_sha256": "4b9e712fcc3ae803ad7f6a66f8242ba3f0c7647b79628eb57e80c4e9b4b877b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "9dbf1216e37e187c3a89f32c1d0aa3e6d00fb350094818fe192d4a6025311685", + "output_sha256": "4b9e712fcc3ae803ad7f6a66f8242ba3f0c7647b79628eb57e80c4e9b4b877b7", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "9dbf1216e37e187c3a89f32c1d0aa3e6d00fb350094818fe192d4a6025311685", + "output_sha256": "cbdfc166f2527bb7cf746316bcd20637c9e6c47453ae77ba026cd8e681399b6b", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "9dbf1216e37e187c3a89f32c1d0aa3e6d00fb350094818fe192d4a6025311685", + "output_sha256": "072036de7a4a8de61742db771a6d2ed0ef6025acc1fb08dbaef8528d3bc92b86", + "pass": false, + "retrieval": true, + "task": "math" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "fddbfa04efed1b45baa60ada446345f59252fc2d93dd0e75c0538b78b671e637", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "fddbfa04efed1b45baa60ada446345f59252fc2d93dd0e75c0538b78b671e637", + "output_sha256": "b7cf3f4f6ec20e1a18cc3d3ac6758ccaeb61f7bb512f7c5a0f164e25968253f5", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "fddbfa04efed1b45baa60ada446345f59252fc2d93dd0e75c0538b78b671e637", + "output_sha256": "b7cf3f4f6ec20e1a18cc3d3ac6758ccaeb61f7bb512f7c5a0f164e25968253f5", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "fddbfa04efed1b45baa60ada446345f59252fc2d93dd0e75c0538b78b671e637", + "output_sha256": "b7cf3f4f6ec20e1a18cc3d3ac6758ccaeb61f7bb512f7c5a0f164e25968253f5", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "fddbfa04efed1b45baa60ada446345f59252fc2d93dd0e75c0538b78b671e637", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "57a351617e3221489e708cca5adee6a4f0f2b3768daad88bd2fb97d5885d3db2", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "57a351617e3221489e708cca5adee6a4f0f2b3768daad88bd2fb97d5885d3db2", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "57a351617e3221489e708cca5adee6a4f0f2b3768daad88bd2fb97d5885d3db2", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "57a351617e3221489e708cca5adee6a4f0f2b3768daad88bd2fb97d5885d3db2", + "output_sha256": "d38e743f3dd9724810ddcbad756bb6dc4c79a12d1e27676b1f7d415e99ae5ea6", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "57a351617e3221489e708cca5adee6a4f0f2b3768daad88bd2fb97d5885d3db2", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "c4f809537c4a2e1a8ac870f81998c10deda8bff2634980d4b1964b8ee0b990db", + "output_sha256": "77365600b1188dd65cd19254b14f469d263c4b22e8f2a88e53c88c4507acc86f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "c4f809537c4a2e1a8ac870f81998c10deda8bff2634980d4b1964b8ee0b990db", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "c4f809537c4a2e1a8ac870f81998c10deda8bff2634980d4b1964b8ee0b990db", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "c4f809537c4a2e1a8ac870f81998c10deda8bff2634980d4b1964b8ee0b990db", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "c4f809537c4a2e1a8ac870f81998c10deda8bff2634980d4b1964b8ee0b990db", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "22b401470577c7d84c7bc1f379354f9c20fb50ccc1981ba54618cc7d2e651967", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "22b401470577c7d84c7bc1f379354f9c20fb50ccc1981ba54618cc7d2e651967", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "22b401470577c7d84c7bc1f379354f9c20fb50ccc1981ba54618cc7d2e651967", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "22b401470577c7d84c7bc1f379354f9c20fb50ccc1981ba54618cc7d2e651967", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "22b401470577c7d84c7bc1f379354f9c20fb50ccc1981ba54618cc7d2e651967", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "40d416f000afeb3b6a95c43329650f821bfd62ce7ffc946ef942275903bef7bc", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "40d416f000afeb3b6a95c43329650f821bfd62ce7ffc946ef942275903bef7bc", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "40d416f000afeb3b6a95c43329650f821bfd62ce7ffc946ef942275903bef7bc", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "40d416f000afeb3b6a95c43329650f821bfd62ce7ffc946ef942275903bef7bc", + "output_sha256": "d38e743f3dd9724810ddcbad756bb6dc4c79a12d1e27676b1f7d415e99ae5ea6", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "40d416f000afeb3b6a95c43329650f821bfd62ce7ffc946ef942275903bef7bc", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "14f169f3ae103d3f4d4c1f3cf832fcfdb1fc58172d561a2d29dc44ea8d00b9b8", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "14f169f3ae103d3f4d4c1f3cf832fcfdb1fc58172d561a2d29dc44ea8d00b9b8", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "14f169f3ae103d3f4d4c1f3cf832fcfdb1fc58172d561a2d29dc44ea8d00b9b8", + "output_sha256": "855c8093349fc5a5178bae1c204aacbaacbe0a2571d4bf0579f05e1b8a1f6f33", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "14f169f3ae103d3f4d4c1f3cf832fcfdb1fc58172d561a2d29dc44ea8d00b9b8", + "output_sha256": "855c8093349fc5a5178bae1c204aacbaacbe0a2571d4bf0579f05e1b8a1f6f33", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "14f169f3ae103d3f4d4c1f3cf832fcfdb1fc58172d561a2d29dc44ea8d00b9b8", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "b6b7cd8eddbe11426d31c818a3b863a27ea9f80bd51b414040084a55cf1ad019", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "b6b7cd8eddbe11426d31c818a3b863a27ea9f80bd51b414040084a55cf1ad019", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "b6b7cd8eddbe11426d31c818a3b863a27ea9f80bd51b414040084a55cf1ad019", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "b6b7cd8eddbe11426d31c818a3b863a27ea9f80bd51b414040084a55cf1ad019", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "b6b7cd8eddbe11426d31c818a3b863a27ea9f80bd51b414040084a55cf1ad019", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 0, + "depth": 0, + "item_id_sha256": "6fab086d4e9a7f9b1c8e8fcda95d924ee608d6740a1e914bde8dc976c30f4c23", + "output_sha256": "5ff85af430e64feb0bdb0965be6b9f60015caa2a85010feb9894ab912c4b365f", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 1, + "depth": 1, + "item_id_sha256": "6fab086d4e9a7f9b1c8e8fcda95d924ee608d6740a1e914bde8dc976c30f4c23", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 2, + "depth": 2, + "item_id_sha256": "6fab086d4e9a7f9b1c8e8fcda95d924ee608d6740a1e914bde8dc976c30f4c23", + "output_sha256": "e0efc59c691f327e594f4f564c7554988efe42a4837f6d2a9ae81b80d523d1dc", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 4, + "depth": 4, + "item_id_sha256": "6fab086d4e9a7f9b1c8e8fcda95d924ee608d6740a1e914bde8dc976c30f4c23", + "output_sha256": "b7cf3f4f6ec20e1a18cc3d3ac6758ccaeb61f7bb512f7c5a0f164e25968253f5", + "pass": true, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "6fab086d4e9a7f9b1c8e8fcda95d924ee608d6740a1e914bde8dc976c30f4c23", + "output_sha256": "84cf40f63827b55c8dc020ffbeed6009473c33677e5f683381dce3a27fd2ac86", + "pass": false, + "retrieval": true, + "task": "arc_challenge" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "b9534d097aea5c641f47519f370f6def9a5c5d3bdf7cbecd41a85d5b510f3119", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "89f24795bf0c48c676d2c9466c4cc41f9130640626729a73dbda87638c499171", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "8814ecacfb27d31254c195907aaa44e13e148afb6966295923741f4aa85849ab", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "31b642e7c63666d2c3bef6df1c352daaef3ffaa9bf0e2efdf95982d0ecf96b9a", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "b3278b8198d8e5299a11a7f685f1d18f52dbad71de4895177c1cb2c944d96286", + "output_sha256": "e0ab8534396b1360519d1c7e6b1270abafc503347bc27705b1d19c46c43cbcfa", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "f41dced4c61766639ab4cd3cb94f76298d07df8f338fc91cb9b65dcece9c04a3", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "c5c80e3e11c697ef8570557a0a4e52005532a7a499f62e406508b23d9501feb5", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "ec6316912107b35fcda5f2cba392ed1c4e2c822168cbe894c5aec8d0b6c73aa1", + "output_sha256": "e770f82df9f292cbdbb5b4113162e57af8db285aff6d421048e5a653caa78311", + "pass": false, + "retrieval": true, + "task": "humaneval" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "43ca9472e2bc4f7bdaf14aae990f60c9611a7db6ea477a3464ddf39f2025baa5", + "output_sha256": "3e186c4f52ce6196672267c72f7bf0a1b3a678db3c3c27de0c46f238063b8d6e", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "081646bf7dda9aaebecedff839be3e6530e4dd9144727afc9788828e550216a7", + "output_sha256": "f568482f36f89bbf7cbe0dbd3c49b07722f7afef90a680a136d40787c6efd369", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "fa8a7be0b128f70e51aa5c3eb960066a95cb882fed6ee907782bcbb47f321e38", + "output_sha256": "80a6f85ea04d5f503b2f90b2c4c5bbf9abb354622facf6ff148f9367182cd6ff", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "3a3d20f37d7290231d39e6c5d5ea5971a88059967ada61a448826aa7faa86821", + "output_sha256": "655197a2a80cd56be921ddb7b0259fad179c4631ba2ffb9054f83468d38b278b", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "0182a06f56abf59b9e4ac5731531134e86e8fc0b644dc19cf4557aee9785fcb7", + "output_sha256": "6eb2adacf75aaf98e01989750edd1c683032f63fa2d1451e4b7052438e55c474", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "d041da94c9c1f6e97345f0acce45783ccd7a63ca8a40c6d420e61ae8d0732685", + "output_sha256": "dc6af96115860c0885723f75fd40735c5255d44189ebc259ef73fd8aabdf0348", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "25afd57d48e7d6a91b36dabb1104efac6f2e98ee6c2e28a2b35f670d46a8fb18", + "output_sha256": "1aae0d5721f27a57893b48a72e93c13e9668ac9a15a602d52e028d47680b3f02", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "3f582f66422402668378f3ffed0128906672027c740c509bcb3e414eb0517a44", + "output_sha256": "d477c6c318d2c8f2edff3272a0003523dbecd8336597f425bf598315de0887a3", + "pass": false, + "retrieval": true, + "task": "mbpp" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "aa1228c12274c5be1e72fedfc0cd9a2141e479594ef0294a927eab89105eae81", + "output_sha256": "fe2470078cec77fe1f7398f580092b7a5a391304a954797c8798fe67eff14265", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "a05c02231e19932d90b962bf49d2db5ce68d02ac3987978b8d142f6c80621755", + "output_sha256": "c7b3a2954602f01e5db8ffc2a7e003db0afb15885c5f1710239957438c7a208c", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "81af7640528bb969968fac7dfc86d049eb06ed8f88368206563f6ff74d583ab6", + "output_sha256": "4ae09930de387985442776a85ecc7981372e69c280d62313766392a452949cf3", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "ea5749a1eec3701fc89425493e2cabf1aaf95a5ef4be9b7c1c71d4c7a9f7bae4", + "output_sha256": "f76f1db9aeaad2e11a9f12cf7858122a47860572110138083401d9d5e2d2ff0c", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "a595e5251c0569a4885a9bb55439d2d7ec69e12be99351381985fa1bc4966c64", + "output_sha256": "9ee8c564e95077fbc8fba84cba512c2e32ee82e4e73c12efa487273e403c90cf", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "bd73f8021b3f22bab8eda0e82a6c2814f16d06b7ffed93e46df2a3f9cfa4808c", + "output_sha256": "fe2470078cec77fe1f7398f580092b7a5a391304a954797c8798fe67eff14265", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "89c8622fd6c328365079ecea806699c1d886e94a947673604c5df2097f4ca3a4", + "output_sha256": "23851d3e4262c494b5e26ed64ff4afb9f780fbc8d122134acefbea750a41f71a", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "df89bdf0a5f2535a71dd3b69407f367101f6491d27ae255b19d65bdf27d8e8d1", + "output_sha256": "9ee8c564e95077fbc8fba84cba512c2e32ee82e4e73c12efa487273e403c90cf", + "pass": false, + "retrieval": true, + "task": "hotpotqa" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "15ba3860a4b29b187d85081f6fcae91117d45ee9c4d23c1d11149bcd766f0297", + "output_sha256": "d478c390615a55e432404b86633920ef1de42bcc85465d7633bbb23317d970d4", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "17f7b99159244221eb0e4c11e526028fd9fbb3cabd48dd3ff07d56a011a4ab09", + "output_sha256": "4ae09930de387985442776a85ecc7981372e69c280d62313766392a452949cf3", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "18b0713943eff9b830cf330c6fe0b07c52967d764a7c8af2d60dbff9da3e6342", + "output_sha256": "7d5ce78e3d8a7c7d0d8c71b620b1bee9d8c5ebc3205e26060516f00ea459e8d6", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "20c00e71c14aea1dc9a8fa2d286133be09f9ae3e5ed0c2d658dc47e403790e2f", + "output_sha256": "6223e40c46ea1d4fb4808986faf64ceb36256c38166af76a4968aae3fc52e8ff", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "819b3414ff474769cbc20d09fb4a9cd087568ac31e913eef38d231e932356273", + "output_sha256": "7e76c49acd016c73e4fc0812f2513b8cf076831e6cf09ecdb7ed252df28cfce2", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "f2610574eb5d89bcd2ed09d391e8dff104928bb7e0e924558ae32f68473ef778", + "output_sha256": "f6c9c9fca4e2d41a37737d4492a433a492593caef3708e6a8605686b0a10efa1", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "5ba9544ac20dc2a134795a03093c8aad8eb29b81e3e72068d88575d8ecf706e1", + "output_sha256": "4ae09930de387985442776a85ecc7981372e69c280d62313766392a452949cf3", + "pass": false, + "retrieval": true, + "task": "2wiki" + }, + { + "applied_depth": 8, + "depth": 8, + "item_id_sha256": "5e96505e0c514d9128ee265e521d058684debc451e461db9d6e87da93af50cdb", + "output_sha256": "020b410f4c729f3f502f955621955c88b30cedbc98a93d9843779231a9db48b1", + "pass": false, + "retrieval": true, + "task": "2wiki" + } + ], + "score_type": "task accuracy/pass@1, never BPB", + "scorer": { + "callable": "score_output", + "module": "jobs.b34_yardstick_credibility_v1", + "raw_generation_modified_before_scoring": false, + "sha256": "cf2018fb3a4a30fd5427806a7f3bedce1e93e9f9735bb06f89fcbefb27a179b7" + }, + "suite": "frozen \u00a77.5 56-item qualification", + "training_graph_deleted_before_open": true + }, + "claim_state": "PAID-A100X4-REASONING-CURRICULUM-MEASURED", + "compression": { + "block_bpb": [ + 1.8091633319854736, + 1.9932173490524292, + 1.8344554901123047, + 1.3260767459869385, + 1.5675348043441772, + 1.5515339374542236, + 1.7852952480316162, + 1.914743423461914, + 1.4587777853012085, + 1.8237273693084717, + 1.427213430404663, + 1.6208724975585938, + 0.9042727947235107, + 2.143134117126465, + 1.161341667175293, + 1.820726990699768, + 1.2604944705963135, + 1.8175941705703735, + 1.0705139636993408, + 1.571967601776123, + 0.9085040092468262, + 1.9233386516571045, + 0.8153854608535767, + 1.9656944274902344, + 2.4340295791625977, + 1.9322422742843628, + 1.3808619976043701, + 1.6026779413223267, + 2.4667270183563232, + 1.973150372505188, + 2.160268783569336, + 2.683443546295166, + 1.3106082677841187, + 2.2027792930603027, + 1.7909464836120605, + 1.89365816116333, + 1.3033673763275146, + 1.2365411520004272, + 1.031322717666626, + 1.7223016023635864, + 1.544898509979248, + 2.379042625427246, + 1.015182614326477, + 1.4566333293914795, + 1.2543933391571045, + 2.0511081218719482, + 1.700197696685791, + 0.9204025268554688 + ], + "by_source": { + "algebraic_stack_permissive": { + "blocks": 12, + "mean_bpb": 1.4679492712020874 + }, + "codeparrot_permissive": { + "blocks": 12, + "mean_bpb": 1.9859495162963867 + }, + "gutenberg_en": { + "blocks": 12, + "mean_bpb": 1.4469140768051147 + }, + "wikipedia_en": { + "blocks": 12, + "mean_bpb": 1.6760507822036743 + } + }, + "mean_bpb": 1.6442159414291382 + }, + "compression_bar": { + "degradation_bpb": -0.034288883209228516, + "final_selection_bpb": 1.5092934370040894, + "initial_selection_bpb": 1.5435823202133179, + "maximum_degradation_bpb": 0.5, + "passed": true + }, + "cost": { + "billed_estimate_usd": 30.313970929672198, + "elapsed_seconds": 10913.029534681991, + "hard_ceiling_usd": 250.0, + "projection_abort_points": [ + 100, + 600 + ] + }, + "created_at": "2026-08-09T12:01:29.628675+00:00", + "failed_bars": [ + "qualification_micro_accuracy_minimum", + "reasoning_D8_correct_minimum", + "reasoning_D8_minus_D1_minimum", + "retrieve_reason_D8_minus_no_retrieval_minimum", + "all_three_reasoning_task_slopes_nonnegative" + ], + "hardware": { + "a100x4_exact": true, + "eight_GPU_forbidden": true, + "gather_validated_on_primary_then_broadcast": true, + "local_world_size": 4, + "maximum_gpu_count": 4, + "passed": true, + "rank_records": [ + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a78414dda2af92a634f0034-hwju2mmn-88340-kv2jl", + "local_rank": 0, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 0, + "selected_device_index": 0, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a78414dda2af92a634f0034-hwju2mmn-88340-kv2jl", + "local_rank": 1, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 1, + "selected_device_index": 1, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a78414dda2af92a634f0034-hwju2mmn-88340-kv2jl", + "local_rank": 2, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 2, + "selected_device_index": 2, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a78414dda2af92a634f0034-hwju2mmn-88340-kv2jl", + "local_rank": 3, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 3, + "selected_device_index": 3, + "visible_cuda_devices": 4, + "world_size": 4 + } + ], + "single_node": true, + "visibility_mode": "all-four-visible-per-rank", + "world_size": 4 + }, + "launch_authorization": { + "GPU_job_namespace": "deepadapt", + "abi": "TRIDENT-OWNER-TEXTUAL-AUTHORIZATION-v1", + "authorization_path": "OWNER-TEXTUAL-AUTHORIZATION-OF-RECORD", + "billing_namespace": "deepadapt", + "created_at": "2026-08-09T08:50:37.240499+00:00", + "decision": "LAUNCH_APPROVED", + "owner_record": { + "abi": "TRIDENT-OWNER-TEXTUAL-AUTHORIZATION-v1", + "authority": "owner-directive-authorization-of-record", + "cryptographic_signature_claimed": false, + "deadline": "2026-08-12T00:58:00+00:00", + "pinned_owner_key_forged_or_reused": false, + "registry_capabilities_sha256": "d76cb6adc92c260b83486839222df37a7a259d9cc81ed00fc6027592fa81ed89", + "registry_key": "owner_binding_spend_authorization_72h_countdown_v1", + "rung_caps_usd": { + "rung1": 50.0, + "rung2": 300.0, + "rung3": 5000.0 + }, + "scheme": "TRIDENT-GIT-RECORDED-OWNER-TEXT-v1", + "source_commit": "2f1b13d15f4aa0e8b05f5deb2936009b7b30ea95", + "t0": "2026-08-09T00:58:00+00:00", + "trident_md_sha256": "72a843a78568dcc6ed42d79ecbe978c62bffc7477de6f02638969c333afc0e7a", + "update_marker": "2026-08-09T01:15:00+00:00" + }, + "owner_record_git_verification": { + "registry_capabilities_sha256": "d76cb6adc92c260b83486839222df37a7a259d9cc81ed00fc6027592fa81ed89", + "source_commit": "2f1b13d15f4aa0e8b05f5deb2936009b7b30ea95", + "trident_md_sha256": "72a843a78568dcc6ed42d79ecbe978c62bffc7477de6f02638969c333afc0e7a" + }, + "package_file_sha256": "b270fd479327fca268505d2550dae016b0f2414c77399c269b4fdece74a16a5b", + "package_hard_abort_usd": 250.0, + "package_payload_sha256": "02f0270eab909fcb18a5fcc87d926371fdb28beb188cf570186f2a94b7c25953", + "pinned_owner_key_forged_or_reused": false, + "private_key_used": false, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "rung1": { + "flavor": "a10g-small", + "gpu_count": 1, + "maximum_spend_usd": 50.0 + }, + "rung2": { + "flavor": "a100x4", + "gpu_count": 4, + "maximum_spend_usd": 300.0, + "updates": 3000 + }, + "rung3": { + "authorized_by_owner_record": true, + "flavor": "h200x4", + "gpu_count": 4, + "maximum_spend_usd": 5000.0, + "not_exercised_by_this_gate_token": true + }, + "verified_authorization_path": "TRIDENT-GIT-RECORDED-OWNER-TEXT-v1" + }, + "package": { + "file_sha256": "b270fd479327fca268505d2550dae016b0f2414c77399c269b4fdece74a16a5b", + "path": "results/b34_task_accuracy_pivot_v1/gate_package.signed.json", + "payload_sha256": "02f0270eab909fcb18a5fcc87d926371fdb28beb188cf570186f2a94b7c25953", + "signature_verified": true + }, + "real_A100x4_audit_smoke": { + "abi": "TRIDENT-B34-REASONING-CURRICULUM-A100X4-GATE-v1-A100X4-AUDIT-SMOKE-RESULT", + "approved_main_run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "audit_run_id": "b34-reasoning-rung2-a100-audit-20260809T085051Z-9a00a202", + "claim_state": "REAL-A100X4-AUDIT-DDP-FIRST-STEP-MEASURED", + "cost": { + "billed_estimate_usd": 1.1077614495500003, + "elapsed_seconds": 398.79412183800014, + "hard_ceiling_usd": 2.0, + "hourly_usd": 10.0, + "timeout_minutes": 12 + }, + "created_at": "2026-08-09T08:58:12.942254+00:00", + "eval_decontamination": { + "algorithm": "lowercase alphanumeric/underscore 13-word SHA-256 shingles", + "base_documents_after": 25623, + "base_documents_before": 25623, + "base_documents_rejected_by_source": { + "algebraic_stack_permissive": 0, + "codeparrot_permissive": 0, + "gutenberg_en": 0, + "wikipedia_en": 0 + }, + "comparator_outputs_loaded": false, + "fingerprints_entered_neural_path": false, + "forbidden_shingle_count": 8822, + "passed": true, + "reasoning_blocks_after": 1, + "reasoning_blocks_before": 1, + "reasoning_blocks_rejected": 0, + "section_7_5_raw_prompts_entered_neural_path": false + }, + "evaluation_artifact_staging": { + "contents_opened_before_training": false, + "manifest": { + "file_sha256": "d3f2da346429a60af19ec470026888a3d4770c337dc70e534eebac4822e35190", + "path": "results/capability_yardstick_v1/dataset_manifest.json" + }, + "neural_path_access": false, + "passed": true, + "preregistration": { + "file_sha256": "e017026c98bc64e5bb05928a3c7dc91dbffa6e6cbd78a3c0528119cf0a4fe970", + "path": "results/capability_yardstick_v1/preregistration.signed.json" + } + }, + "failed_bars": [], + "finite_optimizer_steps": true, + "frozen_yardstick_replay": { + "items": 56, + "manifest": { + "files": [ + { + "bytes": 29505064, + "path": "data/validation-00000-of-00001.parquet", + "repo": "framolfese/2WikiMultihopQA", + "revision": "fe713bfbd1afbca1a65246741a75890405d56a3a", + "rows": 12576, + "sha256": "408e2dbb28edc6c8b9ca3ba0c94d4fc7bf17ffb923766593a3a7f546ab4cba59", + "task": "2wiki" + }, + { + "bytes": 203808, + "path": "ARC-Challenge/test-00000-of-00001.parquet", + "repo": "allenai/ai2_arc", + "revision": "210d026faf9955653af8916fad021475a3f00453", + "rows": 1172, + "sha256": "62f03257e737aed263f55c6abf87c7bb0028a44a6bdd2a26eb1279eb42c1d1e9", + "task": "arc_challenge" + }, + { + "bytes": 419088, + "path": "main/test-00000-of-00001.parquet", + "repo": "openai/gsm8k", + "revision": "740312add88f781978c0658806c59bc2815b9866", + "rows": 1319, + "sha256": "ee7b8da9e381df27b9e3f7758a159ab2bdaa4dbaa910546cbbc47e0cb44e4f59", + "task": "gsm8k" + }, + { + "bytes": 27452575, + "path": "distractor/validation-00000-of-00001.parquet", + "repo": "hotpotqa/hotpot_qa", + "revision": "1908d6afbbead072334abe2965f91bd2709910ab", + "rows": 7405, + "sha256": "c20b638ca82b21d04fe12e14ff417ad05153d4d215a65de54497fca4e972f7c6", + "task": "hotpotqa" + }, + { + "bytes": 83920, + "path": "openai_humaneval/test-00000-of-00001.parquet", + "repo": "openai/openai_humaneval", + "revision": "7dce6050a7d6d172f3cc5c32aa97f52fa1a2e544", + "rows": 164, + "sha256": "2f2871a15fbc95b6c683043359f4ed8e144c5a1c4f24f25f66bc51f598dfcfb6", + "task": "humaneval" + }, + { + "bytes": 353114, + "path": "algebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 1187, + "sha256": "0e04a850dac9cfc758eac0c80afd2d0cc65e29ea72f916c8ac1045b12af937f5", + "task": "math" + }, + { + "bytes": 175032, + "path": "counting_and_probability/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 474, + "sha256": "c324818949c75a6227c0328a8499d58023cbbf6757cfc1736f5ab24cc8d87966", + "task": "math" + }, + { + "bytes": 263835, + "path": "geometry/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 479, + "sha256": "98720f3a8134ad9784fd8951195245264495df03ffc5c9485a9a54ff5ae2431a", + "task": "math" + }, + { + "bytes": 394797, + "path": "intermediate_algebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 903, + "sha256": "e859f72486c951c2017da880541ca729940eaa986b0364eb23e238983df9bfeb", + "task": "math" + }, + { + "bytes": 182003, + "path": "number_theory/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 540, + "sha256": "ba156c9c9fa2040836ba12135fd964e6e67d6bccd8c469ec4a3ad941f8f6fbe9", + "task": "math" + }, + { + "bytes": 267729, + "path": "prealgebra/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 871, + "sha256": "301e5d8ade5c7b5038032532e498bdc9dca799e7f1061a63f2e7d980803240cc", + "task": "math" + }, + { + "bytes": 242241, + "path": "precalculus/test-00000-of-00001.parquet", + "repo": "EleutherAI/hendrycks_math", + "revision": "21a5633873b6a120296cce3e2df9d5550074f4a3", + "rows": 546, + "sha256": "da37474ba3761ae5727af3844a7ba963a735bd6a14c65e6137593669a4fe3b9c", + "task": "math" + }, + { + "bytes": 60864, + "path": "sanitized/test-00000-of-00001.parquet", + "repo": "google-research-datasets/mbpp", + "revision": "4bb6404fdc6cacfda99d4ac4205087b89d32030c", + "rows": 257, + "sha256": "e9e9efa2c0d59ef5e55537a9d126b8f875d5ac010a8d75628d76824884e15850", + "task": "mbpp" + } + ], + "final_test_split_eval_only": true, + "items": 56, + "items_per_task": 8, + "manifest": { + "path": "results/capability_yardstick_v1/dataset_manifest.json", + "sha256": "d3f2da346429a60af19ec470026888a3d4770c337dc70e534eebac4822e35190" + }, + "opened_after_training_graph_deleted": true, + "preregistration": { + "path": "results/capability_yardstick_v1/preregistration.signed.json", + "sha256": "e017026c98bc64e5bb05928a3c7dc91dbffa6e6cbd78a3c0528119cf0a4fe970" + } + }, + "passed": true, + "tasks": [ + "gsm8k", + "math", + "arc_challenge", + "humaneval", + "mbpp", + "hotpotqa", + "2wiki" + ], + "training_graph_deleted_before_open": true, + "used_for_gradient_checkpoint_selection_or_extension": false + }, + "full_OSS_data_staging": { + "all_nonprimary_ranks_reopened_same_cache": true, + "passed": true, + "primary_wrote_one_cache": true, + "rank_private_mkdtemp_used": false, + "shared_workdir": "/tmp/trident-b34-full-data-audit-b34-reasoning-rung2-a100-audit-20260809T085051Z-9a00a202", + "source_count": 4, + "source_names": [ + "algebraic_stack_permissive", + "codeparrot_permissive", + "gutenberg_en", + "wikipedia_en" + ] + }, + "hardware": { + "a100x4_exact": true, + "eight_GPU_forbidden": true, + "gather_validated_on_primary_then_broadcast": true, + "local_world_size": 4, + "maximum_gpu_count": 4, + "passed": true, + "rank_records": [ + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a783f6c3e1f34a7e32c012f-i94xl4wg-5eb69-wjls2", + "local_rank": 0, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 0, + "selected_device_index": 0, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a783f6c3e1f34a7e32c012f-i94xl4wg-5eb69-wjls2", + "local_rank": 1, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 1, + "selected_device_index": 1, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a783f6c3e1f34a7e32c012f-i94xl4wg-5eb69-wjls2", + "local_rank": 2, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 2, + "selected_device_index": 2, + "visible_cuda_devices": 4, + "world_size": 4 + }, + { + "cuda_visible_devices": null, + "device_name": "NVIDIA A100-SXM4-80GB", + "hostname": "j-deepadapt-6a783f6c3e1f34a7e32c012f-i94xl4wg-5eb69-wjls2", + "local_rank": 3, + "local_world_size": 4, + "nvidia_visible_devices": "void", + "rank": 3, + "selected_device_index": 3, + "visible_cuda_devices": 4, + "world_size": 4 + } + ], + "single_node": true, + "visibility_mode": "all-four-visible-per-rank", + "world_size": 4 + }, + "launch_authorization": { + "GPU_job_namespace": "deepadapt", + "abi": "TRIDENT-OWNER-TEXTUAL-AUTHORIZATION-v1", + "authorization_path": "OWNER-TEXTUAL-AUTHORIZATION-OF-RECORD", + "billing_namespace": "deepadapt", + "created_at": "2026-08-09T08:50:37.240499+00:00", + "decision": "LAUNCH_APPROVED", + "owner_record": { + "abi": "TRIDENT-OWNER-TEXTUAL-AUTHORIZATION-v1", + "authority": "owner-directive-authorization-of-record", + "cryptographic_signature_claimed": false, + "deadline": "2026-08-12T00:58:00+00:00", + "pinned_owner_key_forged_or_reused": false, + "registry_capabilities_sha256": "d76cb6adc92c260b83486839222df37a7a259d9cc81ed00fc6027592fa81ed89", + "registry_key": "owner_binding_spend_authorization_72h_countdown_v1", + "rung_caps_usd": { + "rung1": 50.0, + "rung2": 300.0, + "rung3": 5000.0 + }, + "scheme": "TRIDENT-GIT-RECORDED-OWNER-TEXT-v1", + "source_commit": "2f1b13d15f4aa0e8b05f5deb2936009b7b30ea95", + "t0": "2026-08-09T00:58:00+00:00", + "trident_md_sha256": "72a843a78568dcc6ed42d79ecbe978c62bffc7477de6f02638969c333afc0e7a", + "update_marker": "2026-08-09T01:15:00+00:00" + }, + "owner_record_git_verification": { + "registry_capabilities_sha256": "d76cb6adc92c260b83486839222df37a7a259d9cc81ed00fc6027592fa81ed89", + "source_commit": "2f1b13d15f4aa0e8b05f5deb2936009b7b30ea95", + "trident_md_sha256": "72a843a78568dcc6ed42d79ecbe978c62bffc7477de6f02638969c333afc0e7a" + }, + "package_file_sha256": "b270fd479327fca268505d2550dae016b0f2414c77399c269b4fdece74a16a5b", + "package_hard_abort_usd": 250.0, + "package_payload_sha256": "02f0270eab909fcb18a5fcc87d926371fdb28beb188cf570186f2a94b7c25953", + "pinned_owner_key_forged_or_reused": false, + "private_key_used": false, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "rung1": { + "flavor": "a10g-small", + "gpu_count": 1, + "maximum_spend_usd": 50.0 + }, + "rung2": { + "flavor": "a100x4", + "gpu_count": 4, + "maximum_spend_usd": 300.0, + "updates": 3000 + }, + "rung3": { + "authorized_by_owner_record": true, + "flavor": "h200x4", + "gpu_count": 4, + "maximum_spend_usd": 5000.0, + "not_exercised_by_this_gate_token": true + }, + "verified_authorization_path": "TRIDENT-GIT-RECORDED-OWNER-TEXT-v1" + }, + "package": { + "file_sha256": "b270fd479327fca268505d2550dae016b0f2414c77399c269b4fdece74a16a5b", + "path": "results/b34_task_accuracy_pivot_v1/gate_package.signed.json", + "payload_sha256": "02f0270eab909fcb18a5fcc87d926371fdb28beb188cf570186f2a94b7c25953", + "signature_verified": true + }, + "proofwriter": { + "bytes": 15965941, + "license": "Apache-2.0", + "local_sha256": "a37758ca145e681935ec371538b6b36ed97710260651a43ce3825256ed7a84d9", + "path": "data/train-00000-of-00002-6176cf3d78f48858.parquet", + "rendered_block_bytes": 128, + "rendered_block_root_sha256": "8548444c280d20cc9374a9a429a58a0385fb516b4b8605e5e15b7908dc2b78b0", + "repo": "tasksource/proofwriter", + "revision": "ee4d22d49b8486786cbc01c0b6f6f6266e56f3bd", + "rows": 292776, + "section_7_5_or_comparator_data": false, + "selected_rows": 1, + "selection": "lowest SHA256(TRIDENT-B34-REASONING-CURRICULUM-A100X4-GATE-v1|id|question)" + }, + "root": { + "parameters": 927458306, + "sha256": "f18f36417f678f0d47092a4d7d3d26bb85a65eb8a39c557c8dc3b386699eb8a1", + "strict_load_required": true + }, + "section_7_5_data_opened_after_training_graph_deleted": true, + "section_7_5_or_comparator_data_used": false, + "task_accuracy_claimed": false, + "terminal_train_authorized": false, + "train_core_prelude": { + "chunked_delta_parity": { + "chunk": 64, + "enabled": true, + "live_state": { + "deployment_state_detach_unchanged": true, + "maximum_absolute_delta": 0.0, + "passed": true, + "tolerance": 2e-05 + }, + "n_delta_blocks": 1, + "reasoning_curriculum": true, + "worst_rel_err": 5.451602191897109e-07 + }, + "cuda_device_count": 4, + "cuda_peak_memory_bytes": 20615501824, + "device": "cuda:0", + "family": "Core5 parameter-neutral PCOF", + "first_measurement_micro_step_reached": true, + "first_training_step": { + "gradient_ancestry": { + "disconnected": [], + "loss_ancestor": "causal raw-byte codelength terms only", + "nonfinite": [], + "passed": true, + "trainable_tensor_count": 94, + "zero_gradient": [] + }, + "gradient_clip_norm": 0.5, + "gradient_norm_before_clip": 26.847230911254883, + "learning_rate": 1.25e-05, + "loss_bpb": 2.5060086250305176, + "optimizer_step_completed": true, + "training_objective_audit": { + "accuracy_verifier_reward_gradient": false, + "all_targets_are_ordinary_future_bytes": true, + "applied_global_rows": [ + 0 + ], + "event_codelength_bits": [ + 4.295586109161377 + ], + "objective": "uniform raw next-byte L_PC + retrieved OSS reasoning trajectory L_PC + PCOF action/compute rate", + "reasoning_event_weight": 0.25, + "task_or_relation_id_neural_input": false, + "transport_direction_constructed": false, + "update": 1 + }, + "update": 1 + }, + "gate_enable_passed": true, + "geometry": { + "block": 128, + "global_batch": 4, + "updates": 2 + }, + "initial_measurement": { + "block_bpb": [ + 1.554079532623291 + ], + "by_source": { + "a100_audit_smoke": { + "blocks": 1, + "mean_bpb": 1.554079532623291 + } + }, + "mean_bpb": 1.554079532623291 + }, + "last_training_objective_audit": null, + "last_training_step": { + "gradient_ancestry": { + "disconnected": [], + "loss_ancestor": "causal raw-byte codelength terms only", + "nonfinite": [], + "passed": true, + "trainable_tensor_count": 94, + "zero_gradient": [] + }, + "gradient_clip_norm": 0.5, + "gradient_norm_before_clip": 16.25172233581543, + "learning_rate": 2.5e-05, + "loss_bpb": 1.7890183925628662, + "optimizer_step_completed": true, + "training_objective_audit": { + "accuracy_verifier_reward_gradient": false, + "all_targets_are_ordinary_future_bytes": true, + "applied_global_rows": [ + 0 + ], + "event_codelength_bits": [ + 2.909604072570801 + ], + "objective": "uniform raw next-byte L_PC + retrieved OSS reasoning trajectory L_PC + PCOF action/compute rate", + "reasoning_event_weight": 0.25, + "task_or_relation_id_neural_input": false, + "transport_direction_constructed": false, + "update": 2 + }, + "update": 2 + }, + "mode": "real-CUDA full-width train_core prelude smoke", + "parameter_contract": { + "parameters": 927458306, + "tensor_count": 94, + "trainable_parameters": 927458306 + }, + "passed": true, + "post_step_measurement": { + "block_bpb": [ + 1.5892614126205444 + ], + "by_source": { + "a100_audit_smoke": { + "blocks": 1, + "mean_bpb": 1.5892614126205444 + } + }, + "mean_bpb": 1.5892614126205444 + }, + "stability_rows": [ + { + "gradient_ancestry": { + "disconnected": [], + "loss_ancestor": "causal raw-byte codelength terms only", + "nonfinite": [], + "passed": true, + "trainable_tensor_count": 94, + "zero_gradient": [] + }, + "gradient_clip_norm": 0.5, + "gradient_norm_before_clip": 26.847230911254883, + "learning_rate": 1.25e-05, + "loss_bpb": 2.5060086250305176, + "optimizer_step_completed": true, + "training_objective_audit": { + "accuracy_verifier_reward_gradient": false, + "all_targets_are_ordinary_future_bytes": true, + "applied_global_rows": [ + 0 + ], + "event_codelength_bits": [ + 4.295586109161377 + ], + "objective": "uniform raw next-byte L_PC + retrieved OSS reasoning trajectory L_PC + PCOF action/compute rate", + "reasoning_event_weight": 0.25, + "task_or_relation_id_neural_input": false, + "transport_direction_constructed": false, + "update": 1 + }, + "update": 1 + }, + { + "gradient_ancestry": { + "disconnected": [], + "loss_ancestor": "causal raw-byte codelength terms only", + "nonfinite": [], + "passed": true, + "trainable_tensor_count": 94, + "zero_gradient": [] + }, + "gradient_clip_norm": 0.5, + "gradient_norm_before_clip": 16.25172233581543, + "learning_rate": 2.5e-05, + "loss_bpb": 1.7890183925628662, + "optimizer_step_completed": true, + "training_objective_audit": { + "accuracy_verifier_reward_gradient": false, + "all_targets_are_ordinary_future_bytes": true, + "applied_global_rows": [ + 0 + ], + "event_codelength_bits": [ + 2.909604072570801 + ], + "objective": "uniform raw next-byte L_PC + retrieved OSS reasoning trajectory L_PC + PCOF action/compute rate", + "reasoning_event_weight": 0.25, + "task_or_relation_id_neural_input": false, + "transport_direction_constructed": false, + "update": 2 + }, + "update": 2 + } + ] + }, + "verdict": "B34-REASONING-A100X4-AUDIT-DDP-FIRST-STEP-PASS" + }, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "terminal_train_authorized": false, + "training": { + "Core5": { + "batch_chain_sha256": "a0fbab8abe80c60cbbd50beb39ccb2a3308bfe7f4aebce5080248fec6c635cf8", + "chunked_delta_parity": { + "chunk": 64, + "enabled": true, + "live_state": { + "deployment_state_detach_unchanged": true, + "maximum_absolute_delta": 0.0, + "passed": true, + "tolerance": 2e-05 + }, + "n_delta_blocks": 1, + "reasoning_curriculum": true, + "worst_rel_err": 5.451602191897109e-07 + }, + "closure_gain_bpb": 0.034288883209228516, + "consensus": { + "exact_float64_barycenter_then_cast": true, + "passed": true, + "reload_max_abs_delta": 0.0, + "risk_selectors_in_state_dict": [], + "snapshot_count": 3, + "snapshot_updates": [ + 2250, + 2625, + 3000 + ], + "state_file_sha256": "e5d9e2deb5f5796f68a406019e675b398cb9f102ec6b2384fb07c0656153cc97" + }, + "cost_projections": [ + { + "abort_required": false, + "arm": "core", + "elapsed_job_seconds": 721.7677342180104, + "event": "u100_projection", + "hard_projection_usd": 250.0, + "projected_total_seconds": 12613.772705871386, + "projected_total_usd": 35.03825751630941, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 100, + "updates_per_second": 0.29263358098951137 + }, + { + "abort_required": false, + "arm": "core", + "elapsed_job_seconds": 2428.858279142005, + "event": "u600_projection", + "hard_projection_usd": 250.0, + "projected_total_seconds": 12263.041086084748, + "projected_total_usd": 34.06400301690208, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 600, + "updates_per_second": 0.29285605693304534 + } + ], + "family": "Core5", + "final_selection": { + "block_bpb": [ + 1.7654621601104736, + 1.6775509119033813, + 1.7044874429702759, + 1.7253446578979492, + 2.050577163696289, + 2.0643539428710938, + 1.2871246337890625, + 1.4389073848724365, + 1.514498233795166, + 1.541416883468628, + 1.689882755279541, + 1.6241947412490845, + 1.5479305982589722, + 1.7291276454925537, + 0.742131769657135, + 1.6833772659301758, + 0.8771798610687256, + 1.7179515361785889, + 0.9334561228752136, + 2.3756325244903564, + 1.0242202281951904, + 1.7428648471832275, + 0.8020771741867065, + 1.7022837400436401, + 0.8476016521453857, + 1.5991582870483398, + 1.438292384147644, + 1.1722264289855957, + 0.3937585651874542, + 1.7599513530731201, + 1.4755520820617676, + 1.707804560661316, + 1.736224889755249, + 1.3843337297439575, + 1.0880886316299438, + 2.2955803871154785, + 0.7852025032043457, + 1.30742347240448, + 1.615981936454773, + 2.509824752807617, + 1.6650320291519165, + 2.63417387008667, + 2.01503324508667, + 2.276613473892212, + 1.0826815366744995, + 0.7619196176528931, + 1.1244748830795288, + 0.8071067333221436 + ], + "by_source": { + "algebraic_stack_permissive": { + "blocks": 12, + "mean_bpb": 1.5487889051437378 + }, + "codeparrot_permissive": { + "blocks": 12, + "mean_bpb": 1.4082144498825073 + }, + "gutenberg_en": { + "blocks": 12, + "mean_bpb": 1.4065195322036743 + }, + "wikipedia_en": { + "blocks": 12, + "mean_bpb": 1.6736501455307007 + } + }, + "mean_bpb": 1.5092934370040894 + }, + "gradient_ancestry": { + "disconnected": [], + "loss_ancestor": "causal raw-byte codelength terms only", + "nonfinite": [], + "passed": true, + "trainable_tensor_count": 94, + "zero_gradient": [] + }, + "initial_selection": { + "block_bpb": [ + 1.811886191368103, + 1.7424615621566772, + 1.7418839931488037, + 1.7839946746826172, + 2.034852981567383, + 2.0927608013153076, + 1.354263424873352, + 1.4885035753250122, + 1.5473333597183228, + 1.5777952671051025, + 1.72678542137146, + 1.6637890338897705, + 1.5920063257217407, + 1.7698633670806885, + 0.796023964881897, + 1.7139983177185059, + 0.9492695331573486, + 1.7578305006027222, + 0.9827136397361755, + 2.4077413082122803, + 1.0773184299468994, + 1.7754627466201782, + 0.8856061100959778, + 1.7367663383483887, + 0.8759438395500183, + 1.654858112335205, + 1.5155136585235596, + 1.1738884449005127, + 0.4860731065273285, + 1.8243361711502075, + 1.4830073118209839, + 1.7454125881195068, + 1.7178800106048584, + 1.3873536586761475, + 1.0700949430465698, + 2.245384693145752, + 0.8178490400314331, + 1.3257153034210205, + 1.6949796676635742, + 2.4937734603881836, + 1.715647578239441, + 2.6047677993774414, + 1.9932360649108887, + 2.3957927227020264, + 0.8672024607658386, + 0.8868818283081055, + 1.2657018899917603, + 0.8397496938705444 + ], + "by_source": { + "algebraic_stack_permissive": { + "blocks": 12, + "mean_bpb": 1.5751080513000488 + }, + "codeparrot_permissive": { + "blocks": 12, + "mean_bpb": 1.4316455125808716 + }, + "gutenberg_en": { + "blocks": 12, + "mean_bpb": 1.45371675491333 + }, + "wikipedia_en": { + "blocks": 12, + "mean_bpb": 1.7138592004776 + } + }, + "mean_bpb": 1.5435823202133179 + }, + "measured_updates_per_second": 0.293302099513431, + "objective_telemetry": [ + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 1.953125e-07, + "loss_bpb": 1.8657745122909546, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 1 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 1.953125e-06, + "loss_bpb": 1.6200802326202393, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 10 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 1.953125e-05, + "loss_bpb": 1.1269564628601074, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 100 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 1.25e-05, + "loss_bpb": 1.3050168752670288, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 512 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 1.2487810821089253e-05, + "loss_bpb": 1.3557912111282349, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 513 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 8.94427190999916e-06, + "loss_bpb": 1.3663437366485596, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 1000 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 7.3029674334022146e-06, + "loss_bpb": 1.1790846586227417, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 1500 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 7.300534327409847e-06, + "loss_bpb": 1.0677287578582764, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 1501 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 6.324555320336758e-06, + "loss_bpb": 1.2095129489898682, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 2000 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 5.520524474738834e-06, + "loss_bpb": 1.1804962158203125, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 2625 + }, + { + "arm": "core", + "event": "training", + "lambda_effective": 0.0, + "learning_rate": 5.163977794943223e-06, + "loss_bpb": 1.1693590879440308, + "objective": "uniform causal BYTE common prefix", + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 3000 + } + ], + "parameter_contract": { + "parameters": 927458306, + "tensor_count": 94, + "trainable_parameters": 927458306 + }, + "u100_projection": { + "abort_required": false, + "arm": "core", + "elapsed_job_seconds": 721.7677342180104, + "event": "u100_projection", + "hard_projection_usd": 250.0, + "projected_total_seconds": 12613.772705871386, + "projected_total_usd": 35.03825751630941, + "run_id": "b34-reasoning-rung2-20260809T085000Z-c83a0d4e", + "update": 100, + "updates_per_second": 0.29263358098951137 + } + }, + "block": 1024, + "durable_state_before_capability_read": { + "immutable_readback_verified": true, + "path": "language_breakthrough_v1/b34_reasoning_curriculum/b34-reasoning-rung2-20260809T085000Z-c83a0d4e.u3000.state.pt", + "readback_sha256": "e5d9e2deb5f5796f68a406019e675b398cb9f102ec6b2384fb07c0656153cc97", + "repo": "deepadapt/trident-ec", + "revision": "ccdcc7b45e3faff4a35f0455c30e2d6020b95eb4", + "sha256": "e5d9e2deb5f5796f68a406019e675b398cb9f102ec6b2384fb07c0656153cc97" + }, + "eval_decontamination": { + "algorithm": "lowercase alphanumeric/underscore 13-word SHA-256 shingles", + "base_documents_after": 25623, + "base_documents_before": 25623, + "base_documents_rejected_by_source": { + "algebraic_stack_permissive": 0, + "codeparrot_permissive": 0, + "gutenberg_en": 0, + "wikipedia_en": 0 + }, + "comparator_outputs_loaded": false, + "fingerprints_entered_neural_path": false, + "forbidden_shingle_count": 8822, + "passed": true, + "reasoning_blocks_after": 512, + "reasoning_blocks_before": 512, + "reasoning_blocks_rejected": 0, + "section_7_5_raw_prompts_entered_neural_path": false + }, + "global_batch": 20, + "objective": { + "A0_pretraining_audit": { + "accuracy_verifier_reward_comparator_gradient": false, + "answers_are_ordinary_future_bytes_not_privileged_labels": true, + "gradient_terms": [ + "causal raw next-byte codelength", + "causal retrieved raw future-byte codelength", + "PCOF action rate", + "PCOF compute rate" + ], + "new_deployed_parameters": 0, + "task_id_relation_id_operator_id_neural_input": false + }, + "acceptance_is_eval_only": { + "Kimi3_live_same_harness": true, + "Opus5_published_non_equivalent_gold": true, + "depth_grid": [ + 0, + 1, + 2, + 4, + 8 + ], + "no_accuracy_checkpoint_selection_or_training_extension": true, + "primary": "task accuracy/pass@1" + }, + "curriculum": { + "allProofs_field_used": false, + "file": "data/train-00000-of-00002-6176cf3d78f48858.parquet", + "license": "Apache-2.0", + "proof_or_verifier_annotations_in_input_record": false, + "raw_fields": [ + "theory", + "question", + "answer rendered as ordinary UTF-8 bytes" + ], + "revision": "ee4d22d49b8486786cbc01c0b6f6f6266e56f3bd", + "section_7_5_suite_or_Kimi_outputs": false, + "source": "tasksource/proofwriter", + "split": "train" + }, + "equation": "mean L_byte(x_t|x_