free-32cd833d
done · peak unseen_mismatch 79.7
α schedule
α 3.245 ×14 α 1.414 ×14 α 3.9 ×104 α 1.7 ×30 α 3.283 ×16 α 0.9412 ×64 α 3.503 ×13 α 1.551 ×27 α 3.749 ×18 α 3.749 ×100
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 3.749
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 | 55.5 ± 7.4 | 52.0 ± 4.6 | 52.6 ± 5.8 | 50.4 ± 2.9 |
| 110 | 58.0 ± 8.5 | 54.2 ± 6.1 | 53.8 ± 7.3 | 51.2 ± 4.5 |
| 120 | 61.7 ± 10.6 | 57.7 ± 8.5 | 56.8 ± 9.2 | 53.0 ± 6.1 |
| 130 | 65.9 ± 11.8 | 61.7 ± 10.4 | 59.3 ± 10.5 | 55.4 ± 7.6 |
| 140 | 69.9 ± 13.3 | 65.1 ± 11.9 | 63.1 ± 12.7 | 58.1 ± 9.5 |
| 150 | 75.4 ± 14.9 | 69.1 ± 13.3 | 69.1 ± 15.3 | 62.0 ± 11.9 |
| 160 | 80.1 ± 16.0 | 73.6 ± 14.8 | 74.4 ± 17.1 | 67.4 ± 13.8 |
| 170 | 83.8 ± 15.1 | 78.3 ± 14.1 | 77.7 ± 16.7 | 71.5 ± 14.3 |
| 180 | 84.8 ± 13.4 | 79.9 ± 12.6 | 77.0 ± 14.7 | 71.1 ± 12.5 |
| 190 | 88.7 ± 13.7 | 83.8 ± 13.3 | 84.0 ± 15.3 | 77.7 ± 13.8 |
| 200 | 91.0 ± 14.4 | 83.9 ± 15.3 | 88.4 ± 15.6 | 79.6 ± 15.9 |
| 210 | 90.7 ± 16.8 | 77.8 ± 19.0 | 89.2 ± 17.2 | 74.7 ± 18.7 |
| 220 | 89.8 ± 18.3 | 73.4 ± 17.9 | 89.1 ± 18.9 | 71.1 ± 18.2 |
| 230 | 89.4 ± 18.2 | 73.1 ± 18.2 | 88.3 ± 19.6 | 70.8 ± 18.4 |
| 240 | 87.7 ± 18.7 | 71.1 ± 16.9 | 87.0 ± 19.4 | 68.3 ± 16.5 |
| 250 | 88.8 ± 15.6 | 72.1 ± 12.9 | 87.1 ± 17.6 | 67.9 ± 13.5 |
| 260 | 91.9 ± 13.7 | 81.6 ± 13.7 | 90.3 ± 15.2 | 77.8 ± 14.7 |
| 270 | 90.5 ± 14.9 | 74.5 ± 14.4 | 88.9 ± 15.6 | 71.1 ± 14.7 |
| 280 | 90.1 ± 16.4 | 74.2 ± 13.6 | 88.9 ± 16.6 | 73.0 ± 13.5 |
| 290 | 91.5 ± 14.6 | 82.1 ± 11.3 | 89.1 ± 18.5 | 78.7 ± 14.0 |
| 300 | 90.3 ± 15.4 | 78.3 ± 18.2 | 87.5 ± 19.6 | 75.8 ± 18.9 |
| 310 | 82.6 ± 17.4 | 66.5 ± 0.2 | 78.5 ± 21.2 | 62.2 ± 3.9 |
| 320 | 84.2 ± 15.8 | 70.7 ± 3.0 | 77.8 ± 22.2 | 68.4 ± 10.3 |
| 330 | 85.8 ± 14.2 | 81.0 ± 7.2 | 80.1 ± 19.9 | 76.0 ± 13.6 |
| 340 | 86.6 ± 13.2 | 85.5 ± 11.2 | 79.5 ± 20.4 | 79.7 ± 16.4 |
| 350 | 76.7 ± 0.0 | 76.5 ± 0.0 | 62.3 ± 0.0 | 64.5 ± 0.0 |
| 360 | 81.6 ± 0.0 | 80.5 ± 0.0 | 66.1 ± 0.0 | 67.5 ± 0.0 |
| 370 | 81.0 ± 0.0 | 82.0 ± 0.0 | 65.2 ± 0.0 | 68.4 ± 0.0 |
| 380 | 82.3 ± 0.0 | 83.6 ± 0.0 | 66.4 ± 0.0 | 68.2 ± 0.0 |
| 390 | 84.3 ± 0.0 | 84.8 ± 0.0 | 70.2 ± 0.0 | 69.9 ± 0.0 |
| 400 | 84.2 ± 0.0 | 84.1 ± 0.0 | 68.6 ± 0.0 | 70.7 ± 0.0 |