id: a26e3cbaa2f44bde90c0ce900b6efd2b
parent_id: 1f39fc6598af4ae9a80af05a333bebb4
item_type: 1
item_id: ec1518e71453490f8fe5ab75cff7180c
item_updated_time: 1786988380435
title_diff: "[]"
body_diff: "[{\"diffs\":[[0,\"-17 \"],[-1,\"late — CRITICAL CORRECTION to range-bias diagnosis)\\\n\\\n> **Live: G51.** Yesterday's \\\"conditional structure, no scalar fix\\\" conclusion was WRONG — my sweep bin had a combo-enumeration bug (id-ordering check skipped ALL suited/offsuit combos; beat tables were pairs-only). Fixed (91a4fc2), fit chain restarted.\\\n\\\n## Corrected picture (HU, n=43K, v6_mc corpus)\\\n\\\n| Quantity | Value |\\\n|---|---|\\\n| Realized hero-best freq | 0.451 |\\\n| Prior-weighted prediction (t=0) | 0.349 |\\\n| Combo-uniform ATC (t=1) | 0.527 |\\\n| **Temperature match** | **t≈0.55-0.60 → 0.447-0.456, slope 0.86** |\\\n\\\n**Where the underestimation comes from (current best answer): the RangeNet prior is TOO STRONG / TOO CONCENTRATED.** Mass sits on premium types while real holdings include far more weak offsuit/junk. A single ATC-temperature ex\"],[1,\"night — real-logic oracle added)\\\n\\\n> **Live: G51.** Overnight chain: fitting collection → curve MLE fit → oracle-scoring collection (all autonomous).\\\n\\\n## Tonight's additions\\\n- **Real-logic oracle data** (c47d8a4): StrategyBot::act records each bot's OWN chosen action + personality name per (hand, hero, street). With realized holdings revealed, the corpus now supports the EMPIRICAL oracle P(action | holding, config) — real per-config bot logic, re\"],[0,\"pla\"],[1,\"c\"],[0,\"in\"],[-1,\"s most of the mean gap. NOT conditional structure (that was the bug artifact); NOT the corrector (on/off identical — still solid); NOT MC math.\\\n\\\nWhy the prior overconcentrates (hypotheses): softmax 169-class classifier; training labels only from hands that reac\"],[1,\"g the hand-fit proxy curves in the overlap score.\\\n- **User's overlap score, implemented** (fit_likelihood --score): binarized log-score, weight added when the holding would take t\"],[0,\"he\"],[-1,\"d\"],[0,\" observ\"],[-1,\"ation; G48-guided collection distribution narrower than reality.\\\n\\\n## New cheapest experiment (supersedes yesterday's pessimism)\\\n**Range temperature as a production knob**: post-process NN ranges `q = (1-t)·p + t·combo_uniform` at inference (probs_to_hand_range or corrector input), t≈0.55 start. Measured target exists (realized mean/slope per bucket). Full gates required — G48 thresholds are co-tuned to the strong prior; expect decision shifts everywhere.\\\n\\\n## Still running overnight\\\n- Collection equity_v7_fit (~10M) → corrected fit_likelihood chained (bgp_01095181a)\\\n- Fitted curves remain valuable: they condition the temperature on action lines (per-class), potentially better than one global t\\\n\\\n## Morning: check fit_log.txt →\"],[1,\"ed line, subtracted otherwise. v17 prior scores −0.60 (mass on hands that wouldn't play these lines); softening monotonically improves it (t=1 → −0.15). Direction validates: the prior is miscalibrated exactly as the corrected sweep showed.\\\n- Chain: equity_v7_fit collection → fit (corrected binary) → **equity_v8_oracle** collection (own_action data) for the empirical-oracle scoring.\\\n\\\n## Morning plan\\\n1. fit_log.txt → fitted curves + diagnostics\\\n2. Build empirical oracle table from equity_v8_oracle: P(own_action | beat(h), street, size, name-class); re-run overlap score with it as match(h)\\\n3.\"],[0,\" g53\"]],\"start1\":31,\"start2\":31,\"length1\":1819,\"length2\":1247},{\"diffs\":[[0,\"tes:\"],[-1,\" (a)\"],[0,\" fit\"]],\"start1\":1286,\"start2\":1286,\"length1\":12,\"length2\":8},{\"diffs\":[[0,\"rves\"],[-1,\", (b)\"],[1,\" /\"],[0,\" glo\"]],\"start1\":1300,\"start2\":1300,\"length1\":13,\"length2\":10},{\"diffs\":[[0,\"ure \"],[-1,\"t=\"],[1,\"~\"],[0,\"0.55\"],[-1,\", (c) both. Then KK + 7_6 replays\"],[1,\" / oracle-scored variant — full gates (KK, 7_6\"],[0,\", Ha\"]],\"start1\":1322,\"start2\":1322,\"length1\":47,\"length2\":59},{\"diffs\":[[0,\"udit\"],[-1,\".\"],[1,\")\"],[0,\"\\\n\\\n## \"],[-1,\"Lesson recorded\\\nReport sweep mechanics (pair-only tables should have shown as suspiciously narrow supports). The 'impossible' result (actual weaker than ATC) was the tell — an impossible result means measurement bug, not exotic poker truth.\"],[1,\"Measurement state (corrected, 91a4fc2)\\\n- HU: realized hero-best 0.451; prior-weighted 0.349; ATC 0.527; match at t≈0.55-0.60 (slope 0.86)\\\n- Bias = prior overconcentration (RangeNet), not conditional structure (that was the pairs-only enumeration artifact), not the corrector, not MC\\\n\\\n## Key notes\\\nIdeas `f3c42f44` | Roadmap `47031623` | Sanity `d632cea3` | Archive `283e98fe`\"]],\"start1\":1420,\"start2\":1420,\"length1\":250,\"length2\":385}]"
metadata_diff: {"new":{},"deleted":[]}
encryption_cipher_text: 
encryption_applied: 0
updated_time: 2026-08-17T17:45:47.519Z
created_time: 2026-08-17T17:45:47.519Z
is_locked: 0
type_: 13