free-82f9c620
done · peak unseen_mismatch 96.8
α schedule
α 4.7 ×10 α 4.3 ×10 α 4.7 ×10 α 4.8 ×10 α 4.1 ×10 α 4.5 ×10 α 4.3 ×10 α 4.6 ×10 α 4.7 ×10 α 4.3 ×10 α 4.6 ×10 α 4.1 ×10 α 5.0 ×10 α 4.3 ×10 α 4.7 ×10 α 4.9 ×10 α 4.7 ×10 α 1.8 ×10 α 0.2 ×10 α 0.3 ×10 α 0.1 ×10 α 0.4 ×10 α 2.5 ×10 α 4.2 ×10 α 4.1 ×10 α 4.8 ×10 α 1.8 ×10 α 0.1 ×10 α 0.6 ×10 α 0.5 ×110
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 0.5
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 | 60.3 ± 7.5 | 56.7 ± 6.1 | 54.2 ± 5.7 | 51.7 ± 3.4 |
| 110 | 64.9 ± 8.9 | 61.2 ± 7.7 | 56.3 ± 7.5 | 53.6 ± 4.9 |
| 120 | 70.0 ± 9.5 | 66.4 ± 8.4 | 59.5 ± 8.7 | 56.2 ± 6.3 |
| 130 | 74.5 ± 9.2 | 71.3 ± 8.4 | 62.4 ± 9.0 | 59.0 ± 7.0 |
| 140 | 77.2 ± 8.0 | 74.3 ± 7.3 | 63.8 ± 8.9 | 60.6 ± 7.1 |
| 150 | 79.1 ± 6.8 | 76.5 ± 6.1 | 64.5 ± 8.4 | 61.6 ± 6.9 |
| 160 | 80.0 ± 5.5 | 77.5 ± 4.9 | 64.3 ± 7.4 | 61.7 ± 6.4 |
| 170 | 80.3 ± 4.7 | 78.0 ± 4.0 | 64.3 ± 6.8 | 61.7 ± 5.6 |
| 180 | 83.7 ± 5.1 | 81.3 ± 4.6 | 68.1 ± 7.7 | 65.3 ± 6.7 |
| 190 | 88.5 ± 6.5 | 86.2 ± 6.6 | 78.3 ± 10.4 | 74.9 ± 10.1 |
| 200 | 94.8 ± 6.1 | 92.3 ± 7.5 | 91.0 ± 8.7 | 87.7 ± 9.8 |
| 210 | 97.4 ± 4.6 | 93.4 ± 9.3 | 96.1 ± 5.8 | 91.1 ± 10.4 |
| 220 | 99.1 ± 1.3 | 83.8 ± 17.9 | 98.5 ± 2.5 | 81.4 ± 18.6 |
| 230 | 99.2 ± 2.4 | 89.5 ± 9.0 | 99.0 ± 2.9 | 86.4 ± 9.4 |
| 240 | 98.5 ± 1.6 | 94.0 ± 5.3 | 96.8 ± 3.6 | 89.5 ± 7.3 |
| 250 | 92.6 ± 4.5 | 86.2 ± 3.0 | 83.2 ± 11.3 | 73.6 ± 7.1 |
| 260 | 90.1 ± 4.6 | 86.5 ± 4.1 | 80.0 ± 10.0 | 73.8 ± 8.9 |
| 270 | 95.5 ± 2.5 | 90.9 ± 0.8 | 89.7 ± 5.3 | 82.5 ± 1.6 |
| 280 | 99.8 ± 0.3 | 98.4 ± 1.3 | 99.3 ± 0.5 | 96.8 ± 1.4 |