free-91df844a
done · peak unseen_mismatch 93.7
α schedule
α 2.535 ×152 α 1.42 ×148 α 1.42 ×100
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 1.42
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 |
|---|---|---|---|---|
| 100 | 54.3 ± 5.5 | 50.8 ± 2.8 | 52.0 ± 4.7 | 49.7 ± 2.4 |
| 110 | 56.5 ± 7.7 | 51.7 ± 4.5 | 53.8 ± 8.0 | 49.9 ± 3.7 |
| 120 | 60.0 ± 10.4 | 54.0 ± 6.0 | 56.3 ± 10.2 | 51.2 ± 4.9 |
| 130 | 64.7 ± 13.6 | 58.2 ± 9.2 | 60.7 ± 13.3 | 54.6 ± 7.9 |
| 140 | 70.3 ± 15.1 | 63.7 ± 11.4 | 64.7 ± 15.3 | 58.3 ± 10.7 |
| 150 | 74.8 ± 16.6 | 68.8 ± 14.0 | 68.9 ± 17.0 | 62.6 ± 13.1 |
| 160 | 79.1 ± 15.6 | 72.6 ± 14.0 | 73.2 ± 17.0 | 66.8 ± 13.7 |
| 170 | 84.1 ± 16.1 | 77.3 ± 14.5 | 79.9 ± 17.4 | 72.2 ± 14.7 |
| 180 | 87.3 ± 15.2 | 80.5 ± 14.5 | 84.1 ± 16.8 | 76.1 ± 14.9 |
| 190 | 88.8 ± 14.9 | 81.1 ± 14.7 | 85.8 ± 17.0 | 76.9 ± 15.5 |
| 200 | 90.3 ± 12.5 | 80.8 ± 12.6 | 87.1 ± 15.3 | 76.5 ± 13.5 |
| 210 | 93.0 ± 10.0 | 81.8 ± 11.7 | 90.5 ± 12.3 | 77.6 ± 12.9 |
| 220 | 96.2 ± 6.4 | 84.7 ± 12.7 | 94.1 ± 8.8 | 80.9 ± 12.4 |
| 230 | 98.4 ± 3.5 | 84.6 ± 13.3 | 96.9 ± 5.0 | 81.5 ± 12.9 |
| 240 | 99.1 ± 2.0 | 84.6 ± 15.3 | 98.3 ± 3.3 | 82.5 ± 14.5 |
| 250 | 99.8 ± 0.4 | 84.5 ± 16.5 | 99.5 ± 0.9 | 81.9 ± 15.5 |
| 260 | 99.8 ± 0.5 | 83.8 ± 15.8 | 99.6 ± 1.0 | 80.8 ± 16.3 |
| 270 | 100.0 ± 0.0 | 69.4 ± 13.0 | 100.0 ± 0.0 | 67.1 ± 10.8 |
| 280 | 100.0 ± 0.0 | 75.0 ± 19.3 | 100.0 ± 0.0 | 72.5 ± 17.6 |
| 290 | 100.0 ± 0.0 | 68.7 ± 24.4 | 100.0 ± 0.0 | 66.8 ± 23.9 |
| 300 | 100.0 ± 0.0 | 72.4 ± 21.3 | 100.0 ± 0.0 | 72.8 ± 20.2 |
| 310 | 100.0 ± 0.0 | 54.1 ± 0.0 | 100.0 ± 0.0 | 55.7 ± 0.0 |
| 320 | 100.0 ± 0.0 | 69.6 ± 0.0 | 100.0 ± 0.0 | 69.9 ± 0.0 |
| 330 | 100.0 ± 0.0 | 53.1 ± 0.0 | 100.0 ± 0.0 | 54.9 ± 0.0 |
| 340 | 100.0 ± 0.0 | 71.3 ± 0.0 | 100.0 ± 0.0 | 71.9 ± 0.0 |
| 350 | 100.0 ± 0.0 | 74.5 ± 0.0 | 100.0 ± 0.0 | 73.5 ± 0.0 |
| 360 | 100.0 ± 0.0 | 93.8 ± 0.0 | 100.0 ± 0.0 | 93.7 ± 0.0 |