# Round 3 deletion screen: merge 1 Keep incumbent B (`bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae`). **No promotion and no final ranking.** This merge contains all ten supplied results, with each policy kept separate. These are 500-pair screens, not thousands of paired finalist seeds. No simulations were run for this merge. The reserved 900000–903999 block has no evidence here; any final ranking requires measurement items on thousands of paired seeds and the stated finalist promotion rule. Every source citation below refers to an exact input result artifact, not an independently reproduced performance claim. Each row records both supplied locators and its supplied artifact digest; `code_url` points to that result JSON, whose `code` field contains the full strategy. It is not the hash of raw Python. All ten local artifact bytes match the supplied artifact digests; all ten decoded UTF-8 `code` strings match their declared code hashes and parse as Python. The [incumbent JSON](https://api.imd.fun/artifacts/db22aff7167a4798251ed3d20432509f30e511e4958799aba5b17b41bb39b4b6) was fetched and its decoded code hash independently matched B. Direct source diffs confirm the stated guarded deletions; offprice-hold additionally changes its docstring and deletes the now-unused dictionary. Numeric values in the JSON are strings copied from original JSON number tokens (or original strings), preserving source precision without rounding. Booleans remain booleans. Signed differences are candidate minus B. Component field names, additional metric fields and timing objects are preserved, including `signed_retail_diff`/`signed_arb_diff` in below-tick-decay. `complete` is strict assignment completeness: the three explicitly disclosed multi-process runs are false even though their source reports `complete=true` and all 500 measured pairs. `flags.source_complete` preserves that source claim; no simulated failures are invented. Comparator-output inconsistencies are separate flags, not proof that the code comparator hash differs. All rows supply overall and s=1..4 counts and moments: n=500 overall; n=115,116,138,131 by spread. Recomputed mean and sample SE from `sum_d`/`sum_d2` agree within floating arithmetic tolerance; spread counts and moments reconcile to overall. These checks validate aggregate consistency only, not execution or seed-level provenance. Source-reported missing seeds, failures and regime mismatch records are preserved below. Full per-seed logs and runnable harnesses were not supplied for independent replication. The source items report simulator commit `b244530ee2a6247b5f930830eeca9dd950da0e7e`, default ChallengeConfig/ParameterVariance, 2000 steps and serial unsandboxed runs. Their reproduction statements are evidence about their own execution, not merge-verified runtime facts. | Deletion (source) | Target s | Target screening gain | SE | Overall gain | SE | Caveat | |---|---:|---:|---:|---:|---:|---| | [remove-offprice-hold](https://api.imd.fun/artifacts/963d2b2458ed69ccbc647c6942245b2b87ca2482f7246a0d5691cffe9a43f65c) | 2 | -0.14272519037586132 | 0.12824675129390192 | -0.03311224416719983 | 0.029776778290655143 | multiple OS processes | | [remove-penny-bank](https://api.imd.fun/artifacts/900c907645c48a484029e1806adc8ad5f5cd4e146c2fcd209a74ed1ede959ee4) | 3 | -3.61575042538768 | 1.615971016940795 | -0.9979471174069997 | 0.4506803176193992 | multiple OS processes | | [remove-flow-stock](https://api.imd.fun/artifacts/bc3f25f2f19c3d5c63c31d9d0d14014e7252e65984db152299cf09645dc4afa7) | 2 | -0.7194414226448272 | 0.9714340375742122 | -0.1669104100535999 | 0.22503491014830912 | B output mismatch | | [remove-inventory-risk-switch](https://api.imd.fun/artifacts/6fa89247389b1728ec651fbb325cd1476aae9cb0dd464cacc701e7b9a23aa63f) | 3 | 0.0 | 0.0 | 0.0 | 0.0 | source warning | | [remove-below-tick-decay](https://api.imd.fun/artifacts/c5826b927527aca1f479e431d10bfd43af3c6b33c76164ba1c506f9a3536510d) | 3 | -0.4818464406956521 | 0.2063634283162034 | -0.13298961763199998 | 0.05761893764574433 | screen only | | [remove-far-probes](https://api.imd.fun/artifacts/9ce35c8335d535e971e2e9da2f69aca2702ad4a35753c64e3ad2d0a0f9ff0ca4) | 2 | -43.97749759798449 | 5.721405984057485 | -10.202779442732401 | 1.5623014330892397 | screen only | | [remove-near-ceiling-stock](https://api.imd.fun/artifacts/c6de3e1267a61fa349e9a576759134f4f8511fd4bcdc03ae635e7797659c551f) | 3 | 0.1320555914427537 | 0.135085496248487 | 0.03644734323820002 | 0.037279249917990115 | multiple OS processes; B output mismatch | | [remove-below-floor-stock](https://api.imd.fun/artifacts/3a6d8e24b5bc146fdd97888138eab98388d56b63aa80a449119fae7563f0282d) | 2 | 0.008955074257757637 | 0.08366724270396629 | 0.002077577227799772 | 0.019347047988881278 | screen only | | [remove-cash-buffer](https://api.imd.fun/artifacts/3faaf36caca3c740d0b0004de07d9a91bec9c7b15af481ce2dbe681cb3033831) | 3 | -2.843469088978261 | 2.8446376510787283 | -0.784797468558 | 0.7851182965752846 | screen only | | [remove-final-override](https://api.imd.fun/artifacts/b2f173220084ac2e5bdf511d38fd84c48e3049f40a82cbc22a610c70c0a052c7) | 2 | 0.0012887931034482431 | 0.0006427682955674524 | 0.00029899999999999236 | 0.00015060877799124369 | screen only | Spread 1 and spread 4 have exactly zero paired effects in all ten source rows; none of these mechanisms targets those spreads. Spread 2 deletions mostly fail to show a screening benefit: removing far probes is strongly harmful, with both retail and arbitrage components worsening; offprice retention and flow-scaled stock have negative but noisy point estimates, while below-floor stock is approximately neutral. Final-override has a tiny positive estimate; its overall mean does not exceed twice its overall SE. These statements use the corresponding linked rows above, not pooled policy returns. For spread 3, deleting the penny bank and below-tick decay produces negative screening effects; the penny-bank loss is mostly arbitrage. Near-ceiling stock removal has a small positive, noisy estimate with comparator and process caveats. The cash-buffer deletion is noisy and dominated by one loss: source `paired.nonzero_difference_pairs` has seed 410361 at -392.5587498502, against overall sum -392.398734279. Inventory-risk-switch removal is zero on the screen. Its embedded code shows spread-3 kLb=kLf=0.08, so forcing kLb removes no distinction there. Each mechanism and exact change is described in its row's `idea`. Two sources disagree on incumbent outputs despite matching B hashes: flow-stock reports overall B=104.39217582931519 and near-ceiling-stock B=104.39217582931512, whereas the other eight report approximately 104.4137531881468 on the same H block. The discrepant rows also differ in every spread's baseline; their s=4 B is approximately 151.5898790537267 instead of 151.67248160310686. This is much larger than summation rounding. The supplied objects do not establish a cause. Their code comparator hashes match B, but their measured baseline is not cross-replicated; resolve configuration/runtime/input provenance before comparing these rows to the common baseline. No correction or pooling is applied. The requester's supplied source-quality warning names inventory-risk-switch and one source that failed rereading or support, but supplies neither the offending URL nor passage. Its external-source claim is qualified accordingly; no specific citation failure is guessed. Its available code bytes and aggregate arithmetic were verified directly here. Those checks do not independently reproduce the claimed simulations. Offprice-hold, penny-bank and near-ceiling-stock explicitly used respectively three, three and two sequential OS processes. Serial execution avoids parallelism but does not satisfy the assignment's ONE OS process rule. The first two disclose checkpointed chunks without rerunning seeds. Near-ceiling-stock also mentions a separate regime-verification pass without enough detail to count that pass. These deviations remain visible; all sources' complete claims are retained separately from strict merge completeness. Final-override's non-null tuning object explicitly says no tuning; penny-bank discloses behavioral pre-checks outside H and the finalist range. No source reports a positive tuning score. Next measurement should first resolve the two baseline discrepancies and rerun protocol-deviating items in one process if they remain relevant. Far-probe removal is the clearest rejection on screening evidence; retain the probes. The inventory switch is a potential simplification because of equal constants, not evidence of improved edge. Positive noisy stock-deletion estimates and the tiny final-step estimate do not authorize combining policies or promoting them. The promotion ledger is empty and the incumbent code hash remains unchanged. The following single JSON evidence table is authoritative for unrounded values. Rows cite the exact supplied item artifacts; full strategy code remains there. Timing is retained as reported and is not a cross-machine performance ranking. ```json { "incumbent": { "code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "code_url": "https://api.imd.fun/artifacts/db22aff7167a4798251ed3d20432509f30e511e4958799aba5b17b41bb39b4b6", "exact_utf8_hash_verified": true, "promotion": false }, "scope": { "screening_seeds": "410000..410499", "requested_pairs_per_item": "500", "finalist_seeds": "900000..903999", "finalist_evidence_present": false, "pooling_performed": false }, "rows": [ { "key": "remove-offprice-hold", "idea": "Ablation of B, no tuning: in the s==2 branch of manage(), deleted the rule that kept ahead-of-queue orders at the rival quote (comp) when that price is not the selected target t and gain>=1 (and its unused held dict). Target-price retention, including its at-comp extension, and every other branch and parameter are unchanged. The docstring line is the only other edit.", "target_spreads": [ "2" ], "code_sha256": "1cc2c30602c8d9ed916ab9fc7ad49080a3a2019646025b717c661279855de32f", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": false, "failures": [], "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "-16.556122083599913", "sum_d2": "221.76901343691657", "mean_diff": "-0.03311224416719983", "se": "0.029776778290655143", "C_mean_edge": "104.3806409439796", "B_mean_edge": "104.4137531881468", "retail_diff": "-0.07881056528401587", "arb_diff": "0.04569832111599936", "nonzero_pairs": "100" }, "by_spread": { "1": { "n": "115", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "28.495154459731307", "B_mean_edge": "28.495154459731307", "retail_diff": "0.0", "arb_diff": "0.0", "nonzero_pairs": "0" }, "2": { "n": "116", "sum_d": "-16.556122083599913", "sum_d2": "221.76901343691657", "mean_diff": "-0.14272519037586132", "se": "0.12824675129390192", "C_mean_edge": "97.62353728910173", "B_mean_edge": "97.76626247947758", "retail_diff": "-0.33970071243103916", "arb_diff": "0.19697552205172286", "nonzero_pairs": "100" }, "3": { "n": "138", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "128.4054514027384", "B_mean_edge": "128.4054514027384", "retail_diff": "0.0", "arb_diff": "0.0", "nonzero_pairs": "0" }, "4": { "n": "131", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "151.67248160310686", "B_mean_edge": "151.67248160310686", "retail_diff": "0.0", "arb_diff": "0.0", "nonzero_pairs": "0" } } }, "seconds_per_simulation": { "mean": "0.9724066762924194", "first10_pairs": [ { "seed": "410000", "B_sec": "1.2075", "C_sec": "1.6767" }, { "seed": "410001", "B_sec": "0.9828", "C_sec": "1.1118" }, { "seed": "410002", "B_sec": "1.016", "C_sec": "1.3639" }, { "seed": "410003", "B_sec": "1.3858", "C_sec": "1.4649" }, { "seed": "410004", "B_sec": "1.0556", "C_sec": "1.1339" }, { "seed": "410005", "B_sec": "1.6038", "C_sec": "1.0124" }, { "seed": "410006", "B_sec": "1.0338", "C_sec": "1.486" }, { "seed": "410007", "B_sec": "1.0167", "C_sec": "1.3402" }, { "seed": "410008", "B_sec": "0.9896", "C_sec": "0.8568" }, { "seed": "410009", "B_sec": "0.8351", "C_sec": "1.155" } ] }, "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": false, "protocol_violations": [ "ONE_OS_PROCESS violated: three sequential chunks per reproduction." ], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": false, "missing_counts": false, "missing_count_fields": [], "source_warning": false, "notes": [ "Extra docstring edit verified; removal of unused held dictionary accompanies the intended deletion." ], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-offprice-hold/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-offprice-hold", "artifact_sha256": "963d2b2458ed69ccbc647c6942245b2b87ca2482f7246a0d5691cffe9a43f65c", "code_url": "https://api.imd.fun/artifacts/963d2b2458ed69ccbc647c6942245b2b87ca2482f7246a0d5691cffe9a43f65c", "measurement_metadata": { "seeds": "410000..410499" } }, { "key": "remove-penny-bank", "idea": "Ablation of B, no tuning: in spread 3 only, q1 (the tick-1 penny-stock bank demand) is forced to 0.0 by one guarded statement after B's q1 block. Spreads 1, 2, 4 and all other spread-3 branches and parameters are unchanged, including main tick-1 bids, probes, qam and the immediate-cover purchase.", "target_spreads": [ "3" ], "code_sha256": "9b2f3efbe560ac54352f2213e411cb39583d5d5438ace592189743f473896310", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": false, "failures": { "missing_seeds": [], "failed_or_exception_seeds": [], "spread_label_mismatch_seeds": [] }, "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "-498.97355870349986", "sum_d2": "51174.580022606344", "mean_diff": "-0.9979471174069997", "se": "0.4506803176193992", "C_mean_edge": "103.41580607073979", "B_mean_edge": "104.4137531881468", "retail_diff": "-0.12187409271922434", "arb_diff": "-0.8760730246894006", "seeds_with_nonzero_d": "40" }, "by_spread": { "1": { "n": "115", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "28.495154459731307", "B_mean_edge": "28.495154459731307", "retail_diff": "0.0", "arb_diff": "0.0", "seeds_with_nonzero_d": "0" }, "2": { "n": "116", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "97.76626247947758", "B_mean_edge": "97.76626247947758", "retail_diff": "0.0", "arb_diff": "0.0", "seeds_with_nonzero_d": "0" }, "3": { "n": "138", "sum_d": "-498.97355870349986", "sum_d2": "51174.580022606344", "mean_diff": "-3.61575042538768", "se": "1.615971016940795", "C_mean_edge": "124.78970097735072", "B_mean_edge": "128.4054514027384", "retail_diff": "-0.4415727997072736", "arb_diff": "-3.1741776256862337", "seeds_with_nonzero_d": "40" }, "4": { "n": "131", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "151.67248160310686", "B_mean_edge": "151.67248160310686", "retail_diff": "0.0", "arb_diff": "0.0", "seeds_with_nonzero_d": "0" } } }, "seconds_per_simulation": "1.1793247027623002", "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": false, "protocol_violations": [ "ONE_OS_PROCESS violated: three sequential processes (175, 268, 57 pairs), disclosed in process_deviation." ], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": false, "missing_counts": false, "missing_count_fields": [], "source_warning": false, "notes": [ "Non-H behavioral pre-checks on seeds 1..8 and spread-3 seeds in 100..199; source reports no tuning. They do not overlap H or reserved finalists." ], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-penny-bank/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-penny-bank", "artifact_sha256": "900c907645c48a484029e1806adc8ad5f5cd4e146c2fcd209a74ed1ede959ee4", "code_url": "https://api.imd.fun/artifacts/900c907645c48a484029e1806adc8ad5f5cd4e146c2fcd209a74ed1ede959ee4", "measurement_metadata": {} }, { "key": "remove-flow-stock", "idea": "Delete only the observed-retail-flow multiplier of stock_rate in the self.s == 2 branch. Keep flow estimation, all other branches, quote policies and parameters byte-identical. Candidate frozen before the first pair.", "target_spreads": [ "2" ], "code_sha256": "403f0ba663012f8837436bcbc1b07e0f58e2eca8b3816c7802d4ad69fe82480f", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": true, "failures": [], "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "-83.45520502679996", "sum_d2": "12648.786883463794", "mean_diff": "-0.1669104100535999", "se": "0.22503491014830912", "C_mean_edge": "104.2252654192616", "B_mean_edge": "104.39217582931519", "retail_diff": "-0.08727035358420006", "arb_diff": "-0.07964005647019996", "C_mean_retail_edge": "130.854916692621", "B_mean_retail_edge": "130.9421870462052", "C_mean_arb_edge": "-26.629651273355602", "B_mean_arb_edge": "-26.550011216885398" }, "by_spread": { "1": { "n": "115", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "28.495164940043477", "B_mean_edge": "28.495164940043477", "retail_diff": "0.0", "arb_diff": "0.0", "C_mean_retail_edge": "45.3629313391313", "B_mean_retail_edge": "45.3629313391313", "C_mean_arb_edge": "-16.86776639903478", "B_mean_arb_edge": "-16.86776639903478" }, "2": { "n": "116", "sum_d": "-83.45520502679996", "sum_d2": "12648.786883463794", "mean_diff": "-0.7194414226448272", "se": "0.9714340375742122", "C_mean_edge": "97.04684543876036", "B_mean_edge": "97.76628686140518", "retail_diff": "-0.37616531717327606", "arb_diff": "-0.34327610547499987", "C_mean_retail_edge": "131.8040748867845", "B_mean_retail_edge": "132.18024020395777", "C_mean_arb_edge": "-34.75722944803793", "B_mean_arb_edge": "-34.413953342562934" }, "3": { "n": "138", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "128.40565590283623", "B_mean_edge": "128.40565590283623", "retail_diff": "0.0", "arb_diff": "0.0", "C_mean_retail_edge": "154.66101846100653", "B_mean_retail_edge": "154.66101846100653", "C_mean_arb_edge": "-26.255362558194204", "B_mean_arb_edge": "-26.255362558194204" }, "4": { "n": "131", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "151.5898790537267", "B_mean_edge": "151.5898790537267", "retail_diff": "0.0", "arb_diff": "0.0", "C_mean_retail_edge": "179.98647334217173", "B_mean_retail_edge": "179.98647334217173", "C_mean_arb_edge": "-28.396594288439694", "B_mean_arb_edge": "-28.396594288439694" } } }, "seconds_per_simulation": { "overall": "1.1641564993741922", "C": "1.1634446446066722", "B": "1.1648683541417122", "simulation_count": "1000", "elapsed_wall_seconds": "1164.2736403723247", "budget_seconds": "2700" }, "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": true, "protocol_violations": [], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": true, "missing_counts": false, "missing_count_fields": [], "source_warning": false, "notes": [ "Incumbent code digest matches B, but reported B_mean_edge differs from the common H baseline overall and in every spread. Cause unresolved; preserve within-item pairs only." ], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-flow-stock/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-flow-stock", "artifact_sha256": "bc3f25f2f19c3d5c63c31d9d0d14014e7252e65984db152299cf09645dc4afa7", "code_url": "https://api.imd.fun/artifacts/bc3f25f2f19c3d5c63c31d9d0d14014e7252e65984db152299cf09645dc4afa7", "measurement_metadata": { "block": "H screening, not reserved finalists", "seed_start": "410000", "seed_end_inclusive": "410499", "requested_pairs": "500", "attempted_pairs": "500", "difference_definition": "C.total_edge-B.total_edge; retail_diff and arb_diff are signed C-B mean component edges. SE=sample_SD(d)/sqrt(n), including all numeric failed results; no failure dropped.", "spread_label": "s=max(1,((first competitor ask or 100)-(first competitor bid or 0)+1)//2), recorded before first on_step for each instance; checked against sample_config and SimulationResult regime.", "regime_checks": { "checked_pairs": "500", "matched_pairs": "500", "mismatched_seeds": [] }, "missing_seeds": [], "failed_seeds": [], "failed_flags": { "C": "0", "B": "0" }, "unchanged_spreads": { "pairs": "384", "all_edges_exactly_equal": true }, "finalist_seeds_used": false } }, { "key": "remove-inventory-risk-switch", "idea": "Ablate the inventory-dependent low-price bid risk switch in observed spread 3 only; use fixed kLb while retaining the low-price branch and all other parameters.", "target_spreads": [ "3" ], "code_sha256": "7e9672118c34389279d5e8b05cbbc20c66d4831ec6e18d3f077fc8eef21fa97d", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": true, "failures": [], "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "104.4137531881468", "B_mean_edge": "104.4137531881468", "retail_diff": "0.0", "arb_diff": "0.0" }, "by_spread": { "1": { "n": "115", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "28.495154459731307", "B_mean_edge": "28.495154459731307", "retail_diff": "0.0", "arb_diff": "0.0" }, "2": { "n": "116", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "97.76626247947758", "B_mean_edge": "97.76626247947758", "retail_diff": "0.0", "arb_diff": "0.0" }, "3": { "n": "138", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "128.4054514027384", "B_mean_edge": "128.4054514027384", "retail_diff": "0.0", "arb_diff": "0.0" }, "4": { "n": "131", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "151.67248160310686", "B_mean_edge": "151.67248160310686", "retail_diff": "0.0", "arb_diff": "0.0" } } }, "seconds_per_simulation": "0.6161293632127345", "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": true, "protocol_violations": [], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": false, "missing_counts": false, "missing_count_fields": [], "source_warning": true, "notes": [ "Requester supplied one failed/unreadable/unsupported source warning for this item, without the URL or quoted passage. External-source support is qualified. Embedded code hash and deletion independently verified here; supplied code has kLb=kLf=0.08 for spread 3, supporting inertness without that external citation." ], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-inventory-risk-switch/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-inventory-risk-switch", "artifact_sha256": "6fa89247389b1728ec651fbb325cd1476aae9cb0dd464cacc701e7b9a23aa63f", "code_url": "https://api.imd.fun/artifacts/6fa89247389b1728ec651fbb325cd1476aae9cb0dd464cacc701e7b9a23aa63f", "measurement_metadata": { "seed_start": "410000", "seed_end": "410499", "attempted": "500", "last_attempted_seed": "410499", "missing_suffix": [], "regime_mismatches": [] } }, { "key": "remove-below-tick-decay", "idea": "Ablated only the persistent self.tb-based downward phat adjustment at observed spread 3 by guarding that adjustment with self.s != 3. The self.tb update, posterior, probability floor, all parameters, and all other spread branches are unchanged.", "target_spreads": [ "3" ], "code_sha256": "ed3b20dcfab6258512db8f362872e8ac34772a7c845ce10799976cd20e84b518", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": true, "failures": { "missing_seeds": [], "failed_pairs": [], "exceptions": [], "failed_flags": { "B_true": "0", "C_true": "0" } }, "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "-66.49480881599999", "sum_d2": "837.1686420672839", "mean_diff": "-0.13298961763199998", "se": "0.05761893764574433", "c_mean_edge": "104.2807635705147", "b_mean_edge": "104.41375318814676", "signed_retail_diff": "-0.17602736997900026", "signed_arb_diff": "0.043037752346999976" }, "by_spread": { "1": { "n": "115", "sum_d": "0", "sum_d2": "0", "mean_diff": "0", "se": "0", "c_mean_edge": "28.49515445973131", "b_mean_edge": "28.49515445973131", "signed_retail_diff": "0", "signed_arb_diff": "0" }, "2": { "n": "116", "sum_d": "0", "sum_d2": "0", "mean_diff": "0", "se": "0", "c_mean_edge": "97.7662624794776", "b_mean_edge": "97.7662624794776", "signed_retail_diff": "0", "signed_arb_diff": "0" }, "3": { "n": "138", "sum_d": "-66.49480881599999", "sum_d2": "837.1686420672839", "mean_diff": "-0.4818464406956521", "se": "0.2063634283162034", "c_mean_edge": "127.92360496204267", "b_mean_edge": "128.40545140273835", "signed_retail_diff": "-0.6377803260108705", "signed_arb_diff": "0.1559338853152173" }, "4": { "n": "131", "sum_d": "0", "sum_d2": "0", "mean_diff": "0", "se": "0", "c_mean_edge": "151.67248160310692", "b_mean_edge": "151.67248160310692", "signed_retail_diff": "0", "signed_arb_diff": "0" } } }, "seconds_per_simulation": "0.7743652030809317", "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": true, "protocol_violations": [], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": false, "missing_counts": false, "missing_count_fields": [], "source_warning": false, "notes": [], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-below-tick-decay/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-below-tick-decay", "artifact_sha256": "c5826b927527aca1f479e431d10bfd43af3c6b33c76164ba1c506f9a3536510d", "code_url": "https://api.imd.fun/artifacts/c5826b927527aca1f479e431d10bfd43af3c6b33c76164ba1c506f9a3536510d", "measurement_metadata": { "block": { "seed_start": "410000", "seed_end": "410499", "requested_n": "500", "completed_n": "500" }, "difference": "C.total_edge - B.total_edge", "regime_label_verification": { "label": "Strategy self.s after the first observed rival quote", "compared_to": "sampled regime.competitor_spread_ticks", "mismatches": [] } } }, { "key": "remove-far-probes", "idea": "One deletion from B, guarded by self.s == 2: retain only BUY probe at f=floor(phat*100+1e-9) and SELL probe at f+1, nearest probes to the posterior mean in their existing ladders. Existing bounds/free-cash checks, 0.01 quantity, inference and all other branches and parameters unchanged.", "target_spreads": [ "2" ], "code_sha256": "9f84735d058ef989c3aea8c7c49bc3843fa1a273ceb0aac5cbfaf82b8524198b", "incumbent_code_sha256": "bf0d179f3e47d28819f6fa7e0d25548c2ced43374c216d36460b00a26a26ffae", "complete": true, "failures": [], "tuning": null, "paired": { "overall": { "n": "500", "sum_d": "-5101.3897213662", "sum_d2": "661024.4032527782", "mean_diff": "-10.202779442732401", "se": "1.5623014330892397", "C_mean_edge": "94.2109737454144", "B_mean_edge": "104.4137531881468", "retail_diff": "-4.4621039356772005", "arb_diff": "-5.740675507044" }, "by_spread": { "1": { "n": "115", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "28.495154459731307", "B_mean_edge": "28.495154459731307", "retail_diff": "0.0", "arb_diff": "0.0" }, "2": { "n": "116", "sum_d": "-5101.3897213662", "sum_d2": "661024.4032527782", "mean_diff": "-43.97749759798449", "se": "5.721405984057485", "C_mean_edge": "53.7887648814931", "B_mean_edge": "97.76626247947758", "retail_diff": "-19.23320661929828", "arb_diff": "-24.74429097863793" }, "3": { "n": "138", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "128.4054514027384", "B_mean_edge": "128.4054514027384", "retail_diff": "0.0", "arb_diff": "0.0" }, "4": { "n": "131", "sum_d": "0.0", "sum_d2": "0.0", "mean_diff": "0.0", "se": "0.0", "C_mean_edge": "151.67248160310686", "B_mean_edge": "151.67248160310686", "retail_diff": "0.0", "arb_diff": "0.0" } } }, "seconds_per_simulation": { "B": "0.5556066088085063", "C": "0.5383661085739732", "overall": "0.5469863586912398" }, "flags": { "source_complete": true, "measurement_pairs_complete": true, "protocol_compliant_as_reported": true, "protocol_violations": [], "artifact_sha256_verified": true, "code_sha256_verified": true, "code_syntax_valid": true, "incumbent_hash_matches_required": true, "comparator_output_mismatch": false, "missing_counts": false, "missing_count_fields": [], "source_warning": false, "notes": [], "measurement_independently_replicated": false }, "input_result_locator": "artifacts/goal/batches/r3-remove-current/items/remove-far-probes/result.json", "input_read_locator": ".imd/reads/artifacts/item-remove-far-probes", "artifact_sha256": "9ce35c8335d535e971e2e9da2f69aca2702ad4a35753c64e3ad2d0a0f9ff0ca4", "code_url": "https://api.imd.fun/artifacts/9ce35c8335d535e971e2e9da2f69aca2702ad4a35753c64e3ad2d0a0f9ff0ca4", "measurement_metadata": { "seed_start": "410000", "seed_end": "410499", "planned_n": "500", "missing_seeds": [], "spread_label_method": "First StepState: max(1,((ask or 100)-(bid or 0)+1)//2); checked for both strategies against sample_config competitor.spread_ticks and result.regime.", "failed_flags": [], "failure_policy": "Include returned partial-edge results even if failed=True; external exceptions retained, unpaired seeds listed missing. No failures discarded.", "d_definition": "C.total_edge-B.total_edge", "se_definition": "sample_SD(d)/sqrt(n); signed component differences are C-B means", "spread_verification": { "simulations_checked": "1000", "mismatches": "0" } } }, { "key": "remove-near-ceiling-stock", "idea": "Single deletion in B: for s==3 the full-target qam branch (Y