s2_e3_blk
done · peak unseen_mismatch 96.5
α schedule
α 2.0 ×600 α 3.0 ×600
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 3.0
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 |
|---|---|---|---|---|
| 50 | 50.5 ± 1.9 | 49.9 ± 0.5 | 50.1 ± 1.9 | 50.1 ± 1.5 |
| 100 | 51.8 ± 1.7 | 50.8 ± 1.9 | 50.5 ± 1.7 | 50.4 ± 1.9 |
| 150 | 63.7 ± 13.6 | 55.5 ± 7.2 | 62.4 ± 12.6 | 52.6 ± 6.4 |
| 200 | 80.6 ± 21.2 | 67.9 ± 17.0 | 79.9 ± 21.9 | 66.8 ± 15.8 |
| 250 | 96.6 ± 6.5 | 91.5 ± 8.4 | 95.5 ± 8.6 | 88.7 ± 10.9 |
| 300 | 99.3 ± 0.7 | 98.5 ± 0.9 | 98.6 ± 1.6 | 96.5 ± 1.9 |
| 350 | 98.8 ± 0.8 | 97.6 ± 1.5 | 96.5 ± 2.1 | 94.1 ± 3.4 |
| 400 | 98.3 ± 1.0 | 97.5 ± 1.4 | 94.8 ± 2.5 | 92.0 ± 4.4 |
| 450 | 98.0 ± 0.7 | 96.8 ± 1.6 | 93.2 ± 3.1 | 89.7 ± 4.5 |
| 500 | 97.5 ± 0.5 | 96.3 ± 1.0 | 90.5 ± 2.5 | 87.8 ± 2.4 |
| 550 | 96.8 ± 1.0 | 95.8 ± 1.3 | 88.7 ± 2.4 | 85.0 ± 2.8 |
| 600 | 97.0 ± 0.4 | 95.7 ± 1.6 | 89.5 ± 1.7 | 86.5 ± 2.9 |
| 650 | 94.0 ± 0.8 | 92.8 ± 1.6 | 83.0 ± 2.9 | 80.1 ± 3.7 |
| 700 | 93.4 ± 1.3 | 92.3 ± 1.9 | 81.4 ± 3.9 | 78.2 ± 4.1 |
| 750 | 94.0 ± 1.4 | 92.9 ± 1.7 | 81.4 ± 4.4 | 77.6 ± 4.6 |
| 800 | 93.7 ± 0.6 | 92.5 ± 1.4 | 80.8 ± 3.8 | 77.9 ± 4.3 |
| 850 | 93.8 ± 1.4 | 92.5 ± 2.0 | 80.8 ± 5.3 | 77.3 ± 5.6 |
| 900 | 93.6 ± 1.0 | 92.1 ± 2.1 | 80.2 ± 3.7 | 76.8 ± 4.4 |
| 950 | 93.7 ± 1.3 | 92.0 ± 1.9 | 79.7 ± 3.8 | 76.0 ± 4.6 |
| 1000 | 94.2 ± 1.1 | 92.6 ± 2.3 | 81.6 ± 4.7 | 78.1 ± 6.0 |
| 1050 | 94.2 ± 0.7 | 92.8 ± 1.8 | 80.7 ± 3.0 | 77.8 ± 4.7 |
| 1100 | 94.4 ± 1.1 | 92.8 ± 1.6 | 81.7 ± 3.4 | 78.4 ± 4.2 |
| 1150 | 94.5 ± 0.7 | 92.8 ± 2.4 | 81.7 ± 2.5 | 78.1 ± 4.8 |
| 1200 | 94.4 ± 0.9 | 93.1 ± 1.6 | 81.3 ± 3.8 | 78.4 ± 4.9 |