russell@unturf.com 9f528b4ac2 #000049 §7 #23 (wip): nli_shadow_grid.py — sweep {model × candidate-cap k × aggregation × θc × θe} over a should-demote vs should-not-demote pair; manifest gains 2 more alternates (minilm-explicit, deberta-v3-large) for the model science ... First grid (MiniLM, n=1 bench-qa 89 STRICT cells): the §7 #22 'fails the gate' verdict was config-specific (defaults k=6/θc=0.5/θe=0.9) — the grid finds k=1/θc=0.93/θe=0.5 → catch 8/12 falsification-hard (incl. BOTH recombination fixtures hard-003 Mercury + hard-005 Einstein), 0/89 STRICT FP, 0/26 synthetic-legit FP; 18/28 synthetic recombination. So the lexical-candidate approach is NOT a dead end. (k=1 = no haystack; θe=0.5 tighter than the clean-set 0.9.) Full multi-model + n=3 characterization next.		2026-05-12 18:05:30 -04:00
..
__init__.py	#000026 Phase 2: real-shard baseline + lazy concept loading	2026-05-08 13:10:58 -04:00
demo_plot.py	fan-out: witness audit + 5F extractor + function-sampled demo + docs	2026-05-09 13:19:30 -04:00
export_nli_onnx.py	#000049 §7 #22: speedup (batch + cuda auto-detect + ONNX-int8 export) + the gate-item-4 verdict at proper n	2026-05-12 17:21:32 -04:00
fivef_threshold_calibration.py	#000047 — close: delta_aggregator knob on ForkScore (Option D)	2026-05-11 08:27:38 -04:00
harvest_falsification_proposals.py	#000037 follow-through: persist falsification proposals + harvest live stream	2026-05-10 19:40:00 -04:00
nli_shadow_grid.py	#000049 §7 #23 (wip): nli_shadow_grid.py — sweep {model × candidate-cap k × aggregation × θc × θe} over a should-demote vs should-not-demote pair; manifest gains 2 more alternates (minilm-explicit, deberta-v3-large) for the model science	2026-05-12 18:05:30 -04:00
nli_shadow_sweep.py	#000049 §7 #21: candidate-clause restriction in ShadowNLI.check — helps, doesn't close it	2026-05-12 14:34:44 -04:00
phi_alignment_probe.py	ticket #000034 Phase 1a: φ_linear alignment probe + KAT fixture	2026-05-10 10:23:38 -04:00
prometheus_sigma_sweep_dryrun.py	#000037 §22: dry-run swap to sweep_weights() + §15.4 added	2026-05-10 19:22:03 -04:00
real_shard_baseline.py	#000026 Phase 2: real-shard baseline + lazy concept loading	2026-05-08 13:10:58 -04:00
selfmodel_chain_snapshot.py	#000025 §10.11 + §10.13 + §10.14 — close the 5F battery	2026-05-11 07:41:37 -04:00
t3_bound_calculator.py	ticket #000036 Tier-2: dav1d Option B (conservative B1 envelope) applied in v1	2026-05-11 07:06:50 -04:00
textbooks_manifest.py	ticket #000031 follow-ups B-1 + B-2: alias attribution + source-side title author	2026-05-10 09:35:49 -04:00
witness_sweep.py	fan-out: close #000030 · composition fixtures · witness end-to-end	2026-05-09 13:29:59 -04:00
witness_sweep_cron.sh	fan-out: witness automation · #000018 analysis · #000033 pillar VII (Claude)	2026-05-09 14:36:39 -04:00
witness_to_5f.py	fan-out: witness audit + 5F extractor + function-sampled demo + docs	2026-05-09 13:19:30 -04:00