s4_e5_blk
done · peak unseen_mismatch 68.8
α schedule
α 4.0 ×600 α 5.0 ×600
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 5.0
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 | 51.7 ± 1.9 | 49.6 ± 1.0 | 49.9 ± 2.1 | 50.5 ± 1.2 |
| 100 | 58.7 ± 7.3 | 56.1 ± 6.1 | 52.6 ± 4.2 | 51.5 ± 2.9 |
| 150 | 76.9 ± 6.3 | 75.5 ± 5.2 | 65.1 ± 6.5 | 61.0 ± 5.8 |
| 200 | 82.0 ± 2.5 | 80.4 ± 1.7 | 64.6 ± 3.8 | 63.1 ± 3.1 |
| 250 | 83.9 ± 2.0 | 82.2 ± 2.0 | 64.0 ± 4.2 | 62.4 ± 3.2 |
| 300 | 85.9 ± 1.6 | 84.5 ± 2.2 | 67.5 ± 4.9 | 66.6 ± 4.4 |
| 350 | 88.1 ± 1.8 | 86.5 ± 1.9 | 71.1 ± 4.8 | 68.8 ± 4.5 |
| 400 | 87.5 ± 1.5 | 85.7 ± 1.8 | 68.1 ± 3.7 | 65.1 ± 3.1 |
| 450 | 87.7 ± 1.6 | 85.7 ± 2.2 | 68.7 ± 5.9 | 66.0 ± 5.3 |
| 500 | 89.5 ± 1.8 | 87.9 ± 2.4 | 72.6 ± 6.4 | 68.8 ± 5.4 |
| 550 | 88.7 ± 2.1 | 86.8 ± 2.6 | 70.7 ± 5.7 | 66.5 ± 5.7 |
| 600 | 88.7 ± 1.7 | 86.5 ± 2.0 | 70.2 ± 4.4 | 66.4 ± 4.3 |
| 650 | 87.9 ± 1.6 | 85.8 ± 2.3 | 69.7 ± 4.5 | 65.1 ± 3.6 |
| 700 | 88.3 ± 2.0 | 85.6 ± 2.0 | 70.0 ± 4.7 | 64.5 ± 3.7 |
| 750 | 87.7 ± 1.6 | 85.7 ± 1.6 | 69.1 ± 4.2 | 64.4 ± 3.5 |
| 800 | 87.4 ± 1.4 | 85.7 ± 1.5 | 69.5 ± 3.7 | 65.3 ± 2.9 |
| 850 | 88.0 ± 2.1 | 86.1 ± 2.4 | 69.5 ± 5.4 | 65.6 ± 3.7 |
| 900 | 87.7 ± 1.4 | 86.1 ± 2.3 | 69.1 ± 3.2 | 65.0 ± 3.1 |
| 950 | 87.7 ± 2.0 | 86.1 ± 1.2 | 69.7 ± 5.1 | 66.0 ± 3.9 |
| 1000 | 88.0 ± 1.5 | 86.0 ± 2.1 | 69.7 ± 4.0 | 65.9 ± 3.2 |
| 1050 | 88.4 ± 1.4 | 86.5 ± 2.1 | 70.8 ± 3.8 | 66.4 ± 3.4 |
| 1100 | 88.7 ± 1.7 | 86.2 ± 1.9 | 72.0 ± 4.7 | 66.5 ± 4.0 |
| 1150 | 88.1 ± 1.3 | 86.6 ± 1.6 | 69.5 ± 3.7 | 66.0 ± 3.6 |
| 1200 | 88.6 ± 1.1 | 86.5 ± 1.5 | 70.8 ± 4.2 | 67.2 ± 4.2 |