diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-01 10:54:40 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-04-01 10:54:40 -0500 |
| commit | ff91444879a7035c88b2c3c48859f36fb560c660 (patch) | |
| tree | bbbf458d07a4c17c633fabb3a89c7fab35a20d10 /results/toy_lq/sweep_results.json | |
| parent | 6315e18de1b8640ddf4a818c767f3fc14cc5001e (diff) | |
Add confirmatory supplement: T1-T4 from checkpoints (no retraining)
WARNING: All methods (including BP) show near-zero BP hidden gradients (~1e-12-1e-14)
when computed via manual forward with detached hidden states. This is inconsistent with
the earlier first-priority analysis which showed BP at 2.86e-04. Investigation needed.
T1: 40 rows (4 methods × 10 seeds) - full metrics
T2: 800 rows (support sparsity, 5 thresholds × 4 methods × 10 seeds × 4 layers)
T3: 48 rows (gradient norm distributions, 3 seeds × 4 methods × 4 layers)
T4: 100 rows (active-subset Gamma, 5 thresholds × 2 methods × 10 seeds)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Diffstat (limited to 'results/toy_lq/sweep_results.json')
0 files changed, 0 insertions, 0 deletions