faeval.git/experiments/periodic_refit.py, branch master Unnamed repository; edit this file 'description' to name the repository. Add Phase 9B+9C: periodic refit fails, top-down curriculum neutral 2026-03-26T05:07:01+00:00 YurenHao0426 Blackhao0426@gmail.com 2026-03-26T05:07:01+00:00 05ccd23154d1e9d090178b9d4d5f2c821711e784 Phase 9B (periodic refit K=5 R=1 alpha=0.75): 14.0% — Vec starts random, periodic refits insufficient without offline pretraining. Phase 9C (top-down curriculum): last1_vec=30.8%, last2_vec=31.1% vs DFA=31.2%. Near-neutral. Cold-start problem persists even for single-block Vec. Only Phase 9A's offline prefit + blend handoff (+1.5%) works. The key ingredient is offline Vec training on frozen checkpoint features. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> 
Phase 9B (periodic refit K=5 R=1 alpha=0.75): 14.0% — Vec starts random,
periodic refits insufficient without offline pretraining.

Phase 9C (top-down curriculum): last1_vec=30.8%, last2_vec=31.1% vs DFA=31.2%.
Near-neutral. Cold-start problem persists even for single-block Vec.

Only Phase 9A's offline prefit + blend handoff (+1.5%) works.
The key ingredient is offline Vec training on frozen checkpoint features.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>