{ "scope": "Local technical blog evidence only; pure saved JSON/source reading. No new policy, helper, engine, API or upload calls.", "cases": [ { "id": "narrow_vs_broad_gen68", "title": "Selected loss recovery did not generalize", "same_source": true, "narrow": { "games": 15, "candidate_wins": 6, "parent_wins": 2, "recoveries": 6, "lost_wins": 2, "mean_own_delta": 11667.666666666666, "mean_margin_delta": 24157.6, "source_sha256": "e42362287e17aa02fbb59bc75eb0eac2e1639316317a3ba1233ba7dc5e2d1b4d" }, "broad": { "games": 120, "candidate_wins": 21, "parent_wins": 90, "recoveries": 5, "lost_wins": 74, "mean_own_delta": -10412.1, "mean_margin_delta": -16332.775, "source_sha256": "e42362287e17aa02fbb59bc75eb0eac2e1639316317a3ba1233ba7dc5e2d1b4d" }, "evidence": [ "E01", "E02", "E03", "E04" ], "limits": [ "Recorded opponent actions, not responsive/live win probabilities.", "The panels have different selection and overlap; do not plot as sequential deteriorating live skill.", "Both fragile controls already failed on the narrow panel; aggregate gain never passed win-preservation." ] }, { "id": "runtime388_to397", "title": "A winning record could still fail the selected runtime gate", "failed_responsive": { "wins": 6, "draws": 10, "losses": 0, "games": 16, "callback_threshold_seconds": 1.0, "slow_cases": [ { "case": "world6_s0", "own_overruns": [ { "step": 408, "seconds": 1.2235144999576733 } ], "peer_overruns": [ { "step": 408, "seconds": 1.2330183750018477 } ], "evidence": [ "E06" ] }, { "case": "world6_s1", "own_overruns": [ { "step": 408, "seconds": 1.2421561669325456 } ], "peer_overruns": [ { "step": 408, "seconds": 1.231346792075783 } ], "evidence": [ "E07" ] } ] }, "profile": { "total_instrumented_seconds": 3.304, "market_value_calls": 9344, "market_value_cumulative_seconds": 2.613, "route_calls": 3172, "route_cumulative_seconds": 0.488, "scope": "Instrumented diagnostic; not comparable as raw wall runtime." }, "route_equivalence": { "status": "PASS", "cases": 4200, "helper_calls": 8400, "full_games": 0 }, "market_equivalence": { "status": "PASS", "cases": 3056, "market_helper_calls": 6112, "price_helper_calls": 488028, "policy_callbacks": 0 }, "same_saved_prefix": { "source396_step408_seconds": 1.1291907500708476, "source397_step408_seconds": 0.6557437090668827, "callbacks_each": 409, "all_actions_equal388": true }, "new_fresh397": { "games": 16, "worlds": 8, "wins": 6, "losses": 0, "draws": 10, "maximum_callback": 0.7856981669319794, "maximum_peer_callback": 1.2264145420631394, "candidate_callback_overruns": 0, "peer_runtime_status": "UNQUALIFIED_OBSERVED_OVERRUN" }, "known_failed_world_regression": { "games": 2, "fresh": false, "maximum_callback": 0.6967134999576956, "full_both_semantic_paths_equal": true }, "mechanism": [ "Select shortest retirement target permutation by exact Manhattan length, then allocate only the winning command list.", "Visit only receipt-bearing market times and count intervening periodic consumption algebraically; preserve exact per-price-call inventory and sale-before-consumption ordering." ], "evidence": [ "E05", "E06", "E07", "E08", "E09", "E10", "E11", "E12", "E13", "E14", "E15", "E16" ], "limits": [ "Fresh397 worlds differ from388; use the saved prefix/known two for paired runtime evidence.", "Frozen361 peer still exceeded the one-second criterion, disclosed as unqualified. A prospective gate amendment retained candidate strict runtime and reported peer risk.", "No measured local maximum is a hosted latency guarantee.", "Helper equivalence calls and prefix callbacks are not full games." ] }, { "id": "aggregate_gain_lost_wins390", "title": "Positive average margin hid two overturned wins", "overall": { "games": 191, "candidate_wins": 93, "parent_wins": 94, "recoveries": 1, "lost_wins": 2, "mean_own_delta": 21.005235602094242, "mean_margin_delta": 39.35078534031414, "source_sha256": "43779a14fd536164c5510af36c27db8d77c5fbcceb1218cb1a86cc582539cccd" }, "lost_winner_rows": [ { "episode": 114924423, "seat": 0, "submission": 56652541, "run": "research390_integrated_final191_k3", "parent_run": "research362_margin361_full120_k3", "parent_margin": 463.0, "margin": -126.0, "own_cash": 121758.0, "rival_cash": 121884.0, "margin_delta": -589.0, "own_delta": -662.0, "rival_delta": -73.0, "semantic_paths_equal": false, "first_semantic_difference": 606, "semantic_diff_steps": 114, "status": "DONE" }, { "episode": 114952397, "seat": 0, "submission": 56652541, "run": "research390_integrated_final191_k3", "parent_run": "research362_margin361_full120_k3", "parent_margin": 151.0, "margin": -121.0, "own_cash": 67401.0, "rival_cash": 67522.0, "margin_delta": -272.0, "own_delta": 86.0, "rival_delta": 358.0, "semantic_paths_equal": false, "first_semantic_difference": 614, "semantic_diff_steps": 106, "status": "DONE" } ], "evidence": [ "E17", "E18", "E19" ], "limits": [ "The later literal-suffix cargo certificate omitted adaptive future sales; current mass did not prove protected goods would be discarded.", "One recovered loss did not compensate the win-preservation gate for two lost prior winners." ] }, { "id": "archive_seed_jitter_rejoin", "episode": 115748332, "seat": 1, "differing_action_callbacks": [ 377, 381 ], "different_observation_rows": [ 378, 379, 380, 381 ], "observation_rejoin_row": 382, "archive_seed_cost": 20, "original_seed_callback": 377, "archive_seed_callback": 381, "same_final_banks": [ 71683.0, 71391.0 ], "action_witnesses": [ { "callback": 377, "parent_market": [ [ "SELL", "WHEAT", 32 ], [ "BUY_SEED", "CARROT", 1 ] ], "archive_market": [ [ "SELL", "WHEAT", 32 ] ] }, { "callback": 381, "parent_market": [ [ "SELL", "WHEAT", 25 ], [ "BUY_SEED", "WHEAT", 1 ] ], "archive_market": [ [ "SELL", "WHEAT", 25 ], [ "BUY_SEED", "WHEAT", 1 ], [ "BUY_SEED", "CARROT", 1 ] ] } ], "original_full_path_matches": 12, "original_cases": 13, "repeat_full_path_matches": 13, "repeat_cases": 13, "interpretation": "Timing-dependent optional search is plausible, not a proved triggering-helper cause. State observations rejoin at382; the recorded previous action at row382 still differs. Original mismatch and repeat both retained.", "evidence": [ "E20", "E21", "E22", "E23", "E24", "E14" ] }, { "id": "completed_strategic_loss_not_stuck", "title": "A completed loss and three different clocks", "runtime": { "fetched_utc": "2026-09-30T22:58:50.863226+00:00", "source_sha256": "f6b6e081f9c9a0fb509186b424251c2593051d78bc5bc6a58d6de93e44a6f94f", "submission": 56719676, "game_no": 13, "episode": 115999868, "time": "2026-09-30T22:47:45.178832700Z", "own": 75156, "rival": 76382, "margin": -1226, "seat": 0, "opponent_submission": 56719320, "callbacks": 719, "callback_seconds": 63.222255, "maximum_callback": 2.784472, "minimum_overage": 56.51399299999999, "guard_callbacks": [], "traceback_callbacks": [], "steps": 720, "terminal_status": "DONE", "episode_elapsed_seconds": 200.683457, "replay_sha256": "9fc0a1adad2e23e138609986a31e649875da4e82185421d68eae6e9cac80027e", "logs_sha256": "9d078d189890936a3e11c47b73ff3d2752908d6eeca8b559ba260e07395348cc", "scope": "Completed own callbacks/public replay. Cause requires exact local path parity." }, "same361_388_rows": 720, "388_recovery": false, "actual_wages": { "own": 6646, "rival": 6036, "difference": 610, "extra_day25": 233, "extra_day27": 144, "extra_day28": 233 }, "fertilizer_arithmetic": { "actual_doses": 33, "recorded_suffix_extra_carrot_units": 32, "recorded_suffix_extra_wheat_units": 2, "scope": "Offline fixed-action arithmetic, not a played alternative or recoverable-profit certificate." }, "evidence": [ "E25", "E26", "E27", "E28", "E29", "E30", "E31", "E32" ], "limits": [ "Episode elapsed time includes more than own callbacks; callback durations are wall-clock measurements, not CPU attribution.", "No guard or exception was observed; exact local unchanged outcomes support strategic/economic classification, not a timeout-loss explanation.", "Higher wage bill was buying productive tours. Removing labor blindly would also remove real harvest.", "No terminal cargo explanation: actual own terminal product inventory was empty." ] } ], "visual_datasets": { "panel_comparison": { "evidence": [ "E01", "E02" ], "rows": [ { "panel": "Selected difficult 15", "policy": "251", "wins": 2, "games": 15 }, { "panel": "Selected difficult 15", "policy": "332", "wins": 6, "games": 15 }, { "panel": "Chronological 120", "policy": "251", "wins": 90, "games": 120 }, { "panel": "Chronological 120", "policy": "332", "wins": 21, "games": 120 } ] }, "paired_prefix_latency": { "evidence": [ "E06", "E11", "E13" ], "rows": [ { "phase": "Original 388 responsive callback 408", "seconds": 1.2235144999576733 }, { "phase": "396 route optimization; saved prefix", "seconds": 1.1291907500708476 }, { "phase": "397 plus sparse market; saved prefix", "seconds": 0.6557437090668827 } ], "warning": "Original is responsive measurement; optimized values are separate unprofiled prefix runs. Do not label as controlled hardware microbenchmark." }, "loss_clocks": { "evidence": [ "E25" ], "rows": [ { "metric": "Total episode elapsed", "seconds": 200.683457 }, { "metric": "Sum of own callback durations", "seconds": 63.222255 }, { "metric": "Longest own callback", "seconds": 2.784472 } ] } }, "sources": { "E01": { "path": "evaluation/first40_autoresearch_20260928/research332_gen68_funded_labor_current15_k3_review.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/research332_gen68_funded_labor_current15_k3_review.json", "sha256": "62db8930bb798dd54360ff43bd163654f56e09c9c4d85de752e44dafb9e483c0", "locator": "rows, recoveries, lost_wins, source.sha256" }, "E02": { "path": "evaluation/first40_autoresearch_20260928/research333_gen68_full120_k3_review.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/research333_gen68_full120_k3_review.json", "sha256": "229535f2f21841adea820311905c3b631eb47703b8986695a0ac15d34b9fd6c3", "locator": "rows, recoveries, lost_wins, source.sha256" }, "E03": { "path": "evaluation/first40_autoresearch_20260928/lanes333/cohort_summary.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes333/cohort_summary.json", "sha256": "6858a3e2397ece52acf1f136f0f89b64553b02f442a2a1a95e7c65da0f6736f0", "locator": "cohorts" }, "E04": { "path": "evaluation/first40_autoresearch_20260928/REPORT333.md", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/REPORT333.md", "sha256": "9da8175fece9c97d79f3fba3c0cb35c67f41819f5cc8bc2578104f878889b706", "locator": "Coverage and physical limitations" }, "E05": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/completion.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/completion.json", "sha256": "1e074125f07522e8e90613f4a54bca5ec40c85460b0ac091985ccc0c1f755b9b" }, "E06": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/games/world6_s0.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/games/world6_s0.json", "sha256": "12bee48b5865a6bcb2d1996b4c9044780b649984f6fb6f160643598bc3ccb496", "locator": "own_callback_overruns; peer_callback_overruns" }, "E07": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/games/world6_s1.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive388/games/world6_s1.json", "sha256": "7424ad8c0acd2ba6e8d510c0e35fe39d0071179babab1d3d9b1e0422d39bac99", "locator": "own_callback_overruns; peer_callback_overruns" }, "E08": { "path": "evaluation/first40_autoresearch_20260928/lanes396/profile388_original/tottime.txt", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes396/profile388_original/tottime.txt", "sha256": "fc2494835bbcbd410eaa8fb57e6fcda1fa17a6909aeb332d0437876dff393b50" }, "E09": { "path": "evaluation/first40_autoresearch_20260928/lanes396/profile388_original/summary.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes396/profile388_original/summary.json", "sha256": "421e1ddd26edf9dca8baaf458e2c4390d106b17f0ca08447d0de47f2f25592f5", "locator": "policy_callbacks and rows" }, "E10": { "path": "evaluation/first40_autoresearch_20260928/lanes396/route_equivalence.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes396/route_equivalence.json", "sha256": "7c7e2302cdab37421e65662f6e2a25706365f08bdd574c613747f6d97843c136" }, "E11": { "path": "evaluation/first40_autoresearch_20260928/lanes396/route_prefix396.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes396/route_prefix396.json", "sha256": "98544ed46548723ab6cc7c7d2c701f634bea319d998b66b70c40abbeabdebb5e" }, "E12": { "path": "evaluation/first40_autoresearch_20260928/lanes397/market_equivalence.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes397/market_equivalence.json", "sha256": "ce01431245f72baf06c79bbb7d55bb3856c0252667fb493cc541952ba84498f4" }, "E13": { "path": "evaluation/first40_autoresearch_20260928/lanes397/market_prefix397.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes397/market_prefix397.json", "sha256": "3876e09d874e72e1289f94017020552f07d59dc26aa5f3d42c37ebeabed4f8e2" }, "E14": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/final_qualification.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/final_qualification.json", "sha256": "d8c272d1f97369b5df8a24726374310ccb515f78d22d4ed01cf1c8fedbb53704", "locator": "responsive; known_runtime_regression; limitations" }, "E15": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive397/completion.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/responsive397/completion.json", "sha256": "88644e06ead97219cf7bdcf2caff940374fb739cc680de5c08115085b21b18e9" }, "E16": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/runtime_peer_amendment_proposed/approved_prospective.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/runtime_peer_amendment_proposed/approved_prospective.json", "sha256": "9a90c4c7786839f0ecec8b68894158c1381a5605c9f7bd6f86ac7a9d09236adb" }, "E17": { "path": "evaluation/first40_autoresearch_20260928/research390_integrated_final191_k3_review.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/research390_integrated_final191_k3_review.json", "sha256": "04daba5c81e741b19b5d93ddb2e6cff7abd055b84743997e770454529197f0f3" }, "E18": { "path": "evaluation/first40_autoresearch_20260928/lanes390/cohort_summary.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/lanes390/cohort_summary.json", "sha256": "dc02004348bb0a665cfb6a0b73a7ac33c076fe140977851bb75e98e06c4ef4f2" }, "E19": { "path": "evaluation/first40_autoresearch_20260928/REPORT390.md", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/REPORT390.md", "sha256": "fcc86ababe0afdf73eb828091f3dd4502f3eb2cad8bdfb235c4891743f4bdecf" }, "E20": { "path": "evaluation/first40_autoresearch_20260928/submit397_archive13_k3_review.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submit397_archive13_k3_review.json", "sha256": "2d252d61fd032f9a093924b9e18ce0a2e5a2e11bec1d6412d66c49a38d558e93" }, "E21": { "path": "evaluation/first40_autoresearch_20260928/submit397_archive_repeat13_k3_review.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submit397_archive_repeat13_k3_review.json", "sha256": "7a0c56593707664cec35d0ce10d2d60297d727058f6485b7ee102a44375ed825" }, "E22": { "path": "evaluation/first40_autoresearch_20260928/submission_final_20260930/archive_repeat397_predeclared.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_final_20260930/archive_repeat397_predeclared.json", "sha256": "1e12cccf6023ade438b957efda07ffa21755d2bbd3c19bdc17859079d6132a91" }, "E23": { "path": "claude/nine/out/research397_fastretire_full133_k3_dump/115748332.json.gz", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/claude/nine/out/research397_fastretire_full133_k3_dump/115748332.json.gz", "sha256": "41243e40d402d0a2ff00555efcd6a3f64dc89e926778aab6d8c5a3723cb7ad83" }, "E24": { "path": "claude/nine/out/submit397_archive13_k3_dump/115748332.json.gz", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/claude/nine/out/submit397_archive13_k3_dump/115748332.json.gz", "sha256": "1e49a0c0c8f0adbe784d3411fd8542e13ae54c3cbcd6ac5c3aab8672f992d1d1" }, "E25": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/runtime.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/runtime.json", "sha256": "15356a1a6ac2036b7a4f549f34e9047f8f0e346b5639d4cff4d095c93c61db49" }, "E26": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/semantic_parity394_395.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/semantic_parity394_395.json", "sha256": "6e3bd123d0131614d9f72e72cf811cb7b753295a97630c7cbf482cd506d4bd28" }, "E27": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/EXTRA_HIRES.md", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/EXTRA_HIRES.md", "sha256": "68e46619c69086c3ff98ed7f8c75b9b8d34b93dd148b7ea89ed8f45008350f9a" }, "E28": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/extra_fertilizer_routes.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/extra_fertilizer_routes.json", "sha256": "030d231a9af24c7cf952952a306686ca2b714e9a9a79e81b205fbcef7e92913c" }, "E29": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/public_census.json", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/public_census.json", "sha256": "024f408299cc4e9878293c3d47fbd32bea4f8203f6b9a99947683908793a16fc" }, "E30": { "path": "claude/nine/out/research394_margin361_newlive1_k1_dump/115999868.json.gz", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/claude/nine/out/research394_margin361_newlive1_k1_dump/115999868.json.gz", "sha256": "e75421327bb9b711f4308a48d02f57f5320b790b1e897d4a5c57e403ba9811d0" }, "E31": { "path": "claude/nine/out/research395_carrot388_newlive1_k1_dump/115999868.json.gz", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/claude/nine/out/research395_carrot388_newlive1_k1_dump/115999868.json.gz", "sha256": "07229a17c77b6f66c39288f42ab85e38e3882e4aa427d6673ce7e963c528a248" }, "E32": { "path": "evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/FINDINGS.md", "absolute_path": "/Users/subramanya/Desktop/projects/Kaggriculture/evaluation/first40_autoresearch_20260928/submission_next_20260930/loss115999868/astra/FINDINGS.md", "sha256": "69a0762ca1ac8e46a4782b8012b0e3860f26f0de093373315d13b31cca1d095b" } } }