From 533fd8e7877350bda29384f780abae8bf3ead12a Mon Sep 17 00:00:00 2001 From: YurenHao0426 Date: Wed, 22 Jul 2026 05:19:28 -0500 Subject: experiments: freeze normalized C2 recovery --- ROADMAP.md | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'ROADMAP.md') diff --git a/ROADMAP.md b/ROADMAP.md index 75efa60..a28f9af 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -201,6 +201,16 @@ mean d4 validation from `78.633 ± 24.866%` to `81.867 ± 8.779%` and `82.450 ± advances and task seeds 6--8 remain unused. Timescale separation reduces basin variance but does not make plain SGD regression of the moving apical map accurate enough. +One final task-seed-0 regression-mechanism screen is allowed before stopping C2 development. +Normalized LMS divides each local A update by the squared norm of its complete instructional +feature `[c, c outer tanh(h_top)]`, addressing the angle/gain separation in C5 without enlarging +the vectorizer. It fixes the A-only prefix to 100 layerwise-K4 steps, keeps the joint phase at +simultaneous K1/e4, and compares only `eta_A in {0.01, 0.05}` over d1/d4 and model seeds 0--2. +Eligibility is identical to the feedback-first screen: all losses finite, d4 mean at least 90%, at +least 2/3 positive depth gains, and lesion mean at least 2 points with at least 2/3 positive. Among +eligible candidates select the highest d4 mean, preferring `eta_A=0.01` within one point. Failure +closes further tent-map tuning; success permits one frozen confirmation on task seeds 6--8. + ### C3. Causal calibration is genuinely amortized **Status: passed on 2026-07-22.** The frozen five-seed confirmation retained 112.9% of the K16/e4 -- cgit v1.2.3