diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-08 14:51:58 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-08 14:51:58 -0500 |
| commit | 0abfa8e96afffd75cdb6a985603dceb55a284427 (patch) | |
| tree | 5543e407b39c0663eb4b3b6679ec392ae9c9b46a /report/cifar_diagnostics.png | |
| parent | 1c4672e465b9145d71cfddeae4e2ebaeeef2964f (diff) | |
Appendix L: upgrade DFA+pen nudging/drift/trajectory to 3-seed (matches SB/CB)
Ran DFA+pen via cifar_resmlp.py on s123, s456. 3-seed:
- acc 0.3601 ± 0.0014 (matches existing 0.363 from round 11)
- deep cos +0.1518 ± 0.0110 (matches existing +0.155 ± 0.025)
- deep nudge -5e-5 (was single-seed -6e-5)
- train Δloss 0.095 ± 0.007 (was single-seed 0.104)
- w2 drift 18.6 ± 0.5 (matches single-seed 18.8)
- embed drift 94.6 ± 1.4 (matches single-seed 92.7)
Single-seed s42 was within noise of the 3-seed mean across all 4 functional
metrics, so the cos-vs-acc dissociation story is unchanged — Appendix L
now just reports 3-seed for all 3 methods consistently.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Diffstat (limited to 'report/cifar_diagnostics.png')
0 files changed, 0 insertions, 0 deletions