free-f457fa6e
done · peak unseen_mismatch 97.2
α schedule
α 4.6 ×10 α 4.7 ×10 α 4.8 ×10 α 5.0 ×10 α 4.3 ×10 α 4.7 ×10 α 4.6 ×10 α 5.0 ×20 α 4.6 ×10 α 3.7 ×10 α 4.3 ×10 α 4.2 ×20 α 4.7 ×10 α 5.0 ×10 α 4.7 ×10 α 1.6 ×10 α 0.0 ×30 α 0.1 ×10 α 2.7 ×10 α 3.6 ×10 α 3.9 ×10 α 5.0 ×10 α 2.7 ×10 α 1.0 ×10 α 0.7 ×10 α 0.0 ×110
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 0.0
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 | 60.8 ± 7.8 | 57.3 ± 6.1 | 54.1 ± 5.9 | 51.6 ± 3.6 |
| 110 | 65.6 ± 9.1 | 61.8 ± 7.8 | 57.0 ± 7.6 | 53.8 ± 5.3 |
| 120 | 70.6 ± 9.5 | 66.8 ± 8.5 | 59.6 ± 8.7 | 56.2 ± 6.5 |
| 130 | 74.9 ± 8.6 | 71.4 ± 7.9 | 62.1 ± 8.9 | 58.8 ± 6.9 |
| 140 | 77.4 ± 7.5 | 74.5 ± 7.0 | 63.6 ± 8.4 | 60.4 ± 6.5 |
| 150 | 79.1 ± 6.0 | 76.5 ± 5.3 | 64.1 ± 7.6 | 61.4 ± 5.9 |
| 160 | 79.9 ± 4.9 | 77.4 ± 4.5 | 64.5 ± 7.2 | 61.7 ± 5.6 |
| 170 | 80.3 ± 4.3 | 78.1 ± 3.5 | 63.9 ± 6.6 | 61.5 ± 5.2 |
| 180 | 83.8 ± 4.6 | 81.6 ± 4.2 | 68.6 ± 7.5 | 65.7 ± 6.4 |
| 190 | 89.3 ± 6.3 | 86.9 ± 6.4 | 79.5 ± 10.3 | 76.0 ± 9.9 |
| 200 | 94.9 ± 6.5 | 92.6 ± 7.4 | 91.4 ± 8.8 | 88.1 ± 9.6 |
| 210 | 97.6 ± 5.3 | 93.1 ± 10.7 | 96.5 ± 6.7 | 90.4 ± 12.2 |
| 220 | 98.5 ± 4.7 | 73.6 ± 25.1 | 97.9 ± 6.4 | 71.0 ± 25.6 |
| 230 | 99.2 ± 2.6 | 83.2 ± 19.7 | 98.5 ± 3.8 | 80.8 ± 19.2 |
| 240 | 99.2 ± 1.5 | 93.7 ± 10.9 | 98.5 ± 2.8 | 91.6 ± 11.3 |
| 250 | 96.5 ± 3.4 | 89.7 ± 0.9 | 92.9 ± 7.1 | 85.7 ± 1.3 |
| 260 | 94.7 ± 5.0 | 91.9 ± 3.9 | 90.0 ± 9.8 | 88.1 ± 7.5 |
| 270 | 91.1 ± 0.0 | 89.4 ± 0.0 | 79.3 ± 0.0 | 78.4 ± 0.0 |
| 280 | 95.5 ± 0.0 | 94.0 ± 0.0 | 86.1 ± 0.0 | 86.0 ± 0.0 |
| 290 | 99.3 ± 0.0 | 99.2 ± 0.0 | 97.5 ± 0.0 | 97.2 ± 0.0 |