3 Commits

Include renames

Author	SHA1	Message	Date
wassnameandClaudypoo	2f5f93f4fb	diag(#40): Q2 rewrite -- {grad,act}x{cos,dot} scores vs exploited&A>0 label, 3-zone pinning, pairset table ... Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>	2026-06-11 10:13:06 +00:00
wassnameandClaudypoo	70697ff36e	diag(#40): pinning plot splits solve/fail/hack + per-pairset AUROC ranking ... Q4 fix: on-policy "solve" was ~exploited = solves+fails (mostly fails). Split by gt_pass into solve/fail/hack (live: 103 hack / 27 solve / 62 fail). Per-pairset ranking: build v_grad from each heading-prefix subset, re-project the SAME stored live c-grads (no model re-run). Finding: behavior pairs AUROC 0.69 vs all-in-one 0.53; reasoning/opportunity anti-aligned (<0.5) -> mixing dilutes. Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>	2026-06-11 06:16:27 +00:00
wassnameandClaudypoo	979daf84fd	feat(#30): mean+k*std online gate replaces fixed quantile; always-show route cols ... Gate calibration: route by live mean + route_std_mid/route_std_rout * std of the pooled cosine-to-v_grad, not a fixed quantile tail. Self-silences -- only the tail that genuinely exceeds the spread routes, so qmass tracks real separation instead of a forced fraction. The authored absolute band is mis-placed (live pos sits far below the synthetic-hack edge; even synthetic solve out-aligns on-policy hack). tablelog: auroc/rout/routE/keep/resid/qmass cols always shown (nan on vanilla) so arm tables line up. Diagnostics: scripts/diag_pinning.py (4-population calibration view, mean+/-2sd band) and scripts/diag_pinning_refresh.py (proves cosine stats recompute from a tracked v-independent gradient cloud on a v_grad refresh -- exact for k=1, sanity 2.5e-16). Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>	2026-06-11 02:56:07 +00:00