From 3bfa10c06d85c55f796326201fa8fa69be6ef2cf Mon Sep 17 00:00:00 2001 From: Yuren Hao Date: Sat, 18 Jul 2026 17:16:53 -0500 Subject: =?UTF-8?q?RESULT=2045:=20no-floor=E5=88=A4=E5=86=B3=3D=E8=B5=B0?= =?UTF-8?q?=E5=BB=8A=E7=9C=9F=E9=97=AD=E5=90=88(=CE=B2=203e-3=E2=86=921.1e?= =?UTF-8?q?-5=E6=97=A0=E6=90=81=E6=9D=BF,=E5=A4=A9=E8=8A=B1=E6=9D=BF?= =?UTF-8?q?=E5=A1=8C260=C3=97),fail-soft=E5=85=91=E7=8E=B0(gn=E2=89=A40.4?= =?UTF-8?q?=E9=9B=B6spike=E6=9D=83=E9=87=8D=E6=9C=AA=E4=B8=AD=E6=AF=92=20v?= =?UTF-8?q?s=20=E7=A1=AC=E5=BA=95=E7=89=88=E6=85=A2=E7=83=A7);=20plain?= =?UTF-8?q?=E5=85=A8=E7=A8=8B=E5=88=A4=E6=AD=BB=E2=86=92centered;=202x2?= =?UTF-8?q?=E5=85=B3=E9=94=AE=E6=A0=BCc190=E5=9C=A8=E9=A3=9E(cent=E4=BC=B0?= =?UTF-8?q?=E8=AE=A1=E5=99=A8=E6=95=91plain=20s190000=E7=8A=B6=E6=80=81=3F?= =?UTF-8?q?)?= 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 | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) (limited to 'docs') diff --git a/docs/campaign/CASCADE_ABLATION_PLAN.md b/docs/campaign/CASCADE_ABLATION_PLAN.md index 9de77a0..2919da6 100644 --- a/docs/campaign/CASCADE_ABLATION_PLAN.md +++ b/docs/campaign/CASCADE_ABLATION_PLAN.md @@ -568,6 +568,25 @@ direction), not training-under-fault; wave-2 = co-training with faults injected +### RESULT 45 (2026-07-18): NO-FLOOR VERDICT — the corridor TRULY CLOSES for the plain lineage (ceiling falls 260x with no shelf); fail-soft property of pure tracking CONFIRMED. Plain judged DEAD as a full-schedule 72M recipe; centered-rescue 2x2 cell in flight. +fw72m_plain_nf (resume plain s190000, cap_floor 0, exact mirror): COMPLETED 234k, best 3.3838 — +set at ~195k (matches crown-3's pre-death 3.3836) and NEVER improved after: the cap crushed +beta 3e-3 -> 9.3e-4@200k -> 2e-4@208k -> 1e-4@220k -> 1.1e-5@232k, monotone with NO SHELF. +The ceiling is really gone (not a hard-bottom artifact); with res>0.02 persisting at beta=1e-5 +the meter is reading FREE-PHASE contraction loss (state property, beta-independent) — consistent +with the 07-14 wall-2 regression at 72M/FW ~195k. +FAIL-SOFT CONFIRMED: gn <=0.4 / drift 0.012 / zero skips / no spike / weights un-poisoned +(val drifts 3.5 -> 4.1 by SNR-dead noise-walk) — vs crown-3's hard-bottom fail-BURN (gn 29.5, +cos -0.76, drift 0.39, spike). cap_floor 0 is the strictly safer governor policy; the hard +bottom bought crown-3 its extra 0.011 of best (floor-fight learning 196-204k) at the price of +late-run poisoning. +DECISION (user pre-committed): plain cannot hold the 72M/FW full schedule -> centered. Cheapest +form = the decisive 2x2 cell fw72m_plain_c190 IN FLIGHT (centered est on the SAME plain s190000, +cap_floor 0, otherwise identical): centered holds the ceiling -> HYBRID recipe (plain bulk + +centered tail, ~1.13x cost) is live and 2x2 completes as state-rescuable-by-estimator; +centered also collapses -> plain bulk bakes the disease in before 190k -> full-run centered only +(crown stays fw72m_cent 3.3318 at 1.67x). + ### RESULT 44 (2026-07-18): GAP-SOURCE LOCALIZATION — the residual EP gap lives in the TOP-half blocks, split ~evenly attn/ffn; bottom-half EP grads are loss-equivalent to BP. bpmix 5-arm, all from cent s200000 through the 215k endgame frame (shared frame; internal A/B only, absolutes not comparable to crown lineage). BP reference (allbp) 3.4020; all-EP 3.4245 -- cgit v1.2.3