phi4star
done · peak unseen_mismatch 99.2
Referenced by
α schedule
α 5.0 ×25 α 5.0 ×25 α 5.0 ×25 α 1.0634 ×25 α 1.0536 ×25 α 5.0 ×25 α 5.0 ×25 α 5.0 ×25 α 0.9043 ×25 α 0.8215 ×25 α 0.7122 ×25 α 3.084 ×25 α 2.1282 ×25 α 0.1755 ×25
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 0.1755
seen_* probed at reference α = 1.4 · unseen_* = held-out pairings, uniform (α-independent) · 1000 pairs per condition
| iteration | seen_match | seen_mismatch | unseen_match | unseen_mismatch |
|---|---|---|---|---|
| 25 | 51.1 ± 1.5 | 50.1 ± 1.5 | 50.1 ± 1.7 | 50.0 ± 1.6 |
| 50 | 51.6 ± 1.9 | 50.1 ± 1.7 | 50.2 ± 1.5 | 49.9 ± 1.5 |
| 75 | 53.6 ± 3.3 | 51.7 ± 2.4 | 50.9 ± 2.4 | 49.9 ± 1.6 |
| 100 | 57.0 ± 8.2 | 52.5 ± 4.4 | 53.7 ± 7.7 | 50.6 ± 3.5 |
| 125 | 68.2 ± 15.7 | 55.3 ± 9.1 | 65.3 ± 16.0 | 53.4 ± 8.4 |
| 150 | 85.8 ± 14.4 | 75.2 ± 13.9 | 82.2 ± 16.1 | 70.9 ± 14.0 |
| 175 | 91.0 ± 8.5 | 86.2 ± 9.0 | 85.5 ± 11.4 | 79.6 ± 10.9 |
| 200 | 90.6 ± 5.9 | 87.9 ± 5.8 | 83.0 ± 9.4 | 79.0 ± 9.1 |
| 225 | 99.0 ± 3.3 | 97.5 ± 5.1 | 98.3 ± 4.5 | 96.1 ± 6.6 |
| 250 | 99.9 ± 0.4 | 97.8 ± 6.6 | 99.8 ± 0.5 | 97.0 ± 7.2 |
| 275 | 100.0 ± 0.0 | 96.9 ± 9.5 | 100.0 ± 0.1 | 96.3 ± 10.0 |
| 300 | 96.9 ± 2.3 | 94.1 ± 4.3 | 91.5 ± 6.0 | 85.9 ± 7.8 |
| 325 | 98.0 ± 1.1 | 95.7 ± 2.3 | 93.8 ± 3.5 | 88.6 ± 5.2 |
| 350 | 100.0 ± 0.0 | 99.3 ± 5.2 | 100.0 ± 0.0 | 99.2 ± 5.1 |