diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-03-26 00:07:01 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-03-26 00:07:01 -0500 |
| commit | 05ccd23154d1e9d090178b9d4d5f2c821711e784 (patch) | |
| tree | 0ce74409f506df9f7f1717c13b30e79eb5a24f12 /report_explore/MEMO_5B_frozen_vector_transfer.md | |
| parent | ccc6add69553893f6d3f9de4e2010ca8139ba1a6 (diff) | |
Add Phase 9B+9C: periodic refit fails, top-down curriculum neutral
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>
Diffstat (limited to 'report_explore/MEMO_5B_frozen_vector_transfer.md')
0 files changed, 0 insertions, 0 deletions