From 69314ce4dee3b76225a434263e305cbd2cdb04ae Mon Sep 17 00:00:00 2001 From: Yuren Hao Date: Sat, 18 Jul 2026 22:36:45 -0500 Subject: =?UTF-8?q?RESULT=2047+48:=20=E6=B6=A8=E7=9A=84=E9=87=8F=3Dblocks8?= =?UTF-8?q?-11=E5=8D=95=E6=A8=A1=E6=80=81=E5=A2=9E=E7=9B=8A(=CF=81?= =?UTF-8?q?=E5=A4=8D=E5=88=BB=E6=8E=A2=E9=92=88,=E4=B8=A4=E8=A1=80?= =?UTF-8?q?=E7=BB=9F=E5=90=8C=E6=9E=84=E5=9E=8B90%=E8=B4=A8=E9=87=8F,plain?= =?UTF-8?q?=20205-210k=E8=BF=871,cent=E5=85=A8=E7=A8=8B=E5=B9=B3)+b8=20log?= =?UTF-8?q?it=E8=B7=91=E9=A3=9E70=E2=86=9294;=20BBP=E5=AE=A1=E8=AE=A1=3Dfp?= =?UTF-8?q?32=E6=A8=A1=E6=8B=9F=E5=99=A8=E6=97=A0=E5=8A=A0=E6=80=A7?= =?UTF-8?q?=E4=B8=8B=E7=95=8C(a=3D0,overlap=201.000@3e-4)=E2=86=92BBP?= =?UTF-8?q?=E6=98=AF=E7=A1=AC=E4=BB=B6=E8=AE=BE=E8=AE=A1=E6=96=B9=E7=A8=8B?= =?UTF-8?q?;=20=E4=B8=89=E6=8E=A2=E9=92=88=E5=85=A5=E5=BA=93?= 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 | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) (limited to 'docs/campaign') diff --git a/docs/campaign/CASCADE_ABLATION_PLAN.md b/docs/campaign/CASCADE_ABLATION_PLAN.md index d33f7b6..c74a8dd 100644 --- a/docs/campaign/CASCADE_ABLATION_PLAN.md +++ b/docs/campaign/CASCADE_ABLATION_PLAN.md @@ -568,6 +568,34 @@ direction), not training-under-fault; wave-2 = co-training with faults injected +### RESULT 48 (2026-07-18): BBP FLOOR AUDIT — the fp32 SIMULATOR has NO additive detectability floor (a=0 all layers, top-direction overlap 1.000 down to beta=3e-4); BBP becomes a HARDWARE design equation. +probe_bbp @plain s150000, 24 layers (qkv+w2 x12), 6 batches x beta {3e-4,1e-3,3e-3,1e-2}: +additive coefficient a = 0 within fit resolution EVERYWHERE; residual error is beta-independent +(b ~ 1-7e-7 entry-std, negligible vs sigma1(g) 0.002-0.01); u1-overlap 1.000 at all beta +(slight dip at 1e-2, e.g. 0.898 on b0.w2 = finite-beta bias, wall-2 leakage, NOT noise). +=> The empirical floor's benefit never came through the detectability channel (consistent with +the r-sweep multiplicative finding); the corridor closed from the CEILING side only. Wall-1-as- +BBP is real where noise IS additive iid: ANALOG readout. Design equation for T64: +beta*_l = a_HW(sqrt m + sqrt n)/sigma1(g_l), a_HW from ENOB 8.51b / 32uV window numbers -> +quantitative minimum-nudge spec ("big nudge free" upgraded from empirical to closed form). +Muon footnote: sub-threshold beta + msign orthogonalization = full-magnitude noise injection +(why the hard-bottom burn was toxic: gn pinned ~0.3 by normalization, cos -0.76). + +### RESULT 47 (2026-07-18): THE GROWING QUANTITY NAMED — one relaxation mode living in blocks 8-11 (90% of residual mass, BOTH lineages); its gain crosses 1 at ~205-210k in plain, flat in cent. Weight-space and free-state audits both exonerated. +Three-probe chain: (1) specaudit (weights): ALL spectral scales flat-to-falling; survivor cent +carries LARGER vpath/ffn/logit maxima; qk-gain creep +3% both lineages pre-collapse -> weights +exonerated. (2) statej (free-state per-block sigma(J), FD-JVP): chain-product hypothesis +REFUTED — cent@230k product (3.1e16) exceeds plain@death (3.9e16 @195k) and sails; free-state +norms don't discriminate. (3) rhorelax (governor meter replicated offline, K=30, fixed batch): +res0 plain 0.0020->0.0031 (2.1x cent's flat 0.0015) pre-collapse, 0.0093@210k (rho_tail 1.02, +non-contracting), 0.44/divergent@230k; cent converges ~1e-6 at every step through 230k. +MODE PROFILE: 0.15/0.21/0.26/0.28 on blocks 8-11 — identical in both lineages at all steps. +Same structure as RESULT 44's gap localization: finite mode gain = transmission error (CE gap); +gain->1 = corridor death. One object, two symptoms. State-side discriminator: b8 max|logit| +runs away in plain (70->94 by 195k) vs bounded <=80 in cent (statej data). +TARGETED 1.0x CANDIDATE: logit softcap on blocks 8-11 only (~50*tanh(l/50); healthy p99.9 +~30-50 -> cap inactive except on the runaway tail). Test = one s190000 resume arm, 4h. + ### RESULT 46 (2026-07-18): ENDGAME-42 TABLE — BP is lr-robust (reference stands); plain@3e3 honest 3-seed gap +0.042; ride's edge is MID-phase only. Late probes (cent s150000 -> 160k): fix_late 3.4448 vs rv2as_late 3.4464 — tie/fade (mid-stage ride won by 0.004; late it buys nothing). Riding harvests only where floor << ceiling; late-phase -- cgit v1.2.3