diff options
| author | Yuren Hao <yurenh2@illinois.edu> | 2026-07-09 15:41:54 -0500 |
|---|---|---|
| committer | Yuren Hao <yurenh2@illinois.edu> | 2026-07-09 15:41:54 -0500 |
| commit | 63cd19edce4def028f18e092d1af95ff6e0d91f4 (patch) | |
| tree | 579f60cb62756e8312061e59ed2292955ce65591 /scripts/aep_attention.py | |
| parent | 761ac0bae654333490df7de4c6dc50e24bd863f6 (diff) | |
casc_eq_train: exact mode (dtop_every=1) now DEFAULT — bisection showed the v7 dedup cost ~4% CE at tuned lr; fast mode kept as opt-in. K3 verdict: exact equilibrium-EP 2.0481 vs BP 2.053+-0.004 = matched-tuning PARITY
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn
Diffstat (limited to 'scripts/aep_attention.py')
0 files changed, 0 insertions, 0 deletions