diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-08 17:34:41 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-08 17:34:41 -0500 |
| commit | 4aa9d89a28ad9d5a9cfe1ef685a5e100a648b4ed (patch) | |
| tree | 5bd2aa3c39a0cb9e7c95d6bf6ccbb5760571dc07 /models/residual_mlp.py | |
| parent | 673d78d03f5447fa010083512d20072beacd302b (diff) | |
paper v2.30.2: §3 ¶1 endpoint values use 3-seed means
Paragraph framing says "3 seeds" but the endpoint values were
actually s42-specific. Re-aggregated from
results/snapshot_evolution_v2/snapshot_evolution_s{42,123,456}.json:
||h_L|| final per seed: 4.39e8, 3.86e8, 7.30e8 → mean 5.18e8 (paper now: 5e8, was 4e8)
||g_L|| final per seed: 4.86e-10, 3.76e-10, 2.67e-10 → mean 3.76e-10 (paper now: 4e-10, was 5e-10)
The §5 ¶1 intervention paragraph still says "‖h_L‖ ~4.4e8" — that
one is explicitly the s42 vanilla DFA endpoint that Figure 3 panel (a)
plots, so it stays single-seed. §3 is the 3-seed version.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Diffstat (limited to 'models/residual_mlp.py')
0 files changed, 0 insertions, 0 deletions