From 27560970e2d967e62d22b8693811d7301fbc290d Mon Sep 17 00:00:00 2001 From: Yuren Hao Date: Fri, 24 Jul 2026 18:52:12 -0500 Subject: =?UTF-8?q?RESULT=2060:=20s345000=E8=B5=B0=E5=BB=8A=E7=A9=BA(end07?= =?UTF-8?q?5=E6=9B=B4=E7=B3=9F3.4805,LR=E5=AB=8C=E7=96=91=E6=8E=92?= =?UTF-8?q?=E9=99=A4)=E2=86=92=E9=97=AD=E5=90=88=E7=82=B9275-300k,s250000?= =?UTF-8?q?=3D=E6=9C=80=E5=90=8E=E5=81=A5=E5=BA=B7=E7=82=B9;=20--res=5Fgat?= =?UTF-8?q?e=E5=AE=BD=E5=BA=A6=E5=8F=AF=E9=85=8D+bsign=C3=97beta=5Fsimple?= =?UTF-8?q?=E7=AC=A6=E5=8F=B7=E6=8A=B9=E9=99=A4bug=E4=BF=AE=E5=A4=8D(COMBO?= =?UTF-8?q?=5FCERT=5FPASS);=20fw135m=5Fratchet=E4=BB=8Es250000=E5=8F=91?= =?UTF-8?q?=E5=B0=84(bsign+=E6=A3=98=E8=BD=AE+res=5Fgate=201e-3,27h)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Co-Authored-By: Claude Fable 5 Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn --- docs/campaign/CASCADE_ABLATION_PLAN.md | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'docs') diff --git a/docs/campaign/CASCADE_ABLATION_PLAN.md b/docs/campaign/CASCADE_ABLATION_PLAN.md index a0c8efb..5bb1d73 100644 --- a/docs/campaign/CASCADE_ABLATION_PLAN.md +++ b/docs/campaign/CASCADE_ABLATION_PLAN.md @@ -568,6 +568,28 @@ direction), not training-under-fault; wave-2 = co-training with faults injected +### RESULT 60 (2026-07-24): s345000 CORRIDOR IS EMPTY (end075 3.4805 < original, worse); +closure located at ~275-300k; CONTINUATION = bsign+ratchet+res_gate from s250000. Two code +fixes en route (sign-wipe combo bug + configurable gate), both fire-certified. +end075 (beta 7.5e-4 from s345000): best 3.4805, val climbing to 3.9, gn doubled — LOWER beta +did WORSE than the original 3e-3 over the same segment. LR-scheduler suspect CLEARED (resume +fast-forward exists, line 813). Diagnosis: at s345000 the corridor is EMPTY — ceiling ~1e-3 +(contraction) while the state's elevated relax-noise floor (resK stuck 2.6e-4 at ALL beta, +100x above healthy) pushes the effective lower bound UP past it. Over-ceiling operation bakes +in state damage (the 72M simple-lineage lesson, width edition). +CLOSURE PROBE (resK-to-floor reading, rho-noise verdicts discounted): s250000 CLEAN at 3e-3 +(resK 1.3e-6, ceiling ~5e-3) | s300000 boundary (3e-3 resK 2.2e-5) | s345000 dead. Crossing +~275-300k. Last healthy resume point = s250000. +FIXES: (1) --res_gate flag (0.02 was a C512 constant; C768 ran 100k steps semi-converged under +it; per-width rule ~100x healthy K=3 floor -> 1e-3 for C768); (2) bsign+beta_simple SIGN-WIPE +bug — the beta_simple override sat after the sign flip and silently disabled bsign when both +were on (never manifested: 72M ratchet ran plain). Flip moved after the override. +COMBO_CERT_PASS: forced-knock smoke shows sign flips + knocks + persistent halving together. +IN FLIGHT: fw135m_ratchet — resume s250000, bsign + halve-only ratchet + res_gate 1e-3, +beta0 3e-3, 3-GPU, 190k-step replay (~27h). Expected: knocks track the sinking ceiling +(~290k, ~1.5-2e-3, endgame ~1e-3), anneal harvested inside the corridor. Verdict = best vs +3.3839 (bsign fixed-3e-3) and vs 72M band 3.364. + ### RESULT 59b (2026-07-24): C768 CEILING MEASURED — the run spent its second half 2-3x ABOVE the ceiling, and the ceiling sinks UNDER BSIGN at this width. R59's "graze" and "stability transfers" both CORRECTED. -- cgit v1.2.3