s5_e1_blk
done · peak unseen_mismatch 95.7
α schedule
α 5.0 ×240 α 4.0 ×240 α 3.0 ×240 α 2.0 ×240 α 1.0 ×240
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 1.0
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 | 51.8 ± 2.1 | 49.9 ± 1.2 | 50.2 ± 1.8 | 50.0 ± 1.4 |
| 100 | 59.6 ± 6.4 | 56.9 ± 4.6 | 52.8 ± 3.7 | 51.7 ± 3.0 |
| 150 | 76.5 ± 5.5 | 74.3 ± 5.6 | 62.2 ± 5.9 | 59.1 ± 4.4 |
| 200 | 79.5 ± 1.9 | 78.0 ± 2.0 | 60.5 ± 3.5 | 59.9 ± 3.0 |
| 250 | 80.4 ± 2.0 | 79.1 ± 1.4 | 61.1 ± 3.5 | 59.2 ± 3.5 |
| 300 | 84.8 ± 1.6 | 83.3 ± 1.7 | 65.6 ± 4.9 | 64.4 ± 4.4 |
| 350 | 86.7 ± 2.3 | 85.3 ± 2.5 | 69.1 ± 5.2 | 67.1 ± 5.5 |
| 400 | 86.6 ± 1.2 | 85.4 ± 2.5 | 66.3 ± 4.0 | 63.7 ± 5.0 |
| 450 | 87.8 ± 1.3 | 86.1 ± 1.9 | 68.8 ± 3.9 | 64.7 ± 4.1 |
| 500 | 89.2 ± 2.0 | 87.0 ± 2.3 | 70.8 ± 4.3 | 66.4 ± 4.8 |
| 550 | 90.6 ± 1.6 | 88.4 ± 2.0 | 72.8 ± 3.9 | 67.4 ± 4.4 |
| 600 | 91.5 ± 1.3 | 89.1 ± 1.5 | 73.4 ± 5.2 | 68.3 ± 6.0 |
| 650 | 92.3 ± 1.5 | 89.8 ± 1.8 | 76.3 ± 4.0 | 69.7 ± 5.0 |
| 700 | 93.0 ± 1.0 | 91.3 ± 0.8 | 77.1 ± 3.1 | 72.0 ± 3.7 |
| 750 | 97.7 ± 0.8 | 96.1 ± 1.3 | 90.2 ± 4.0 | 83.0 ± 6.4 |
| 800 | 97.5 ± 0.9 | 95.9 ± 1.2 | 88.6 ± 3.0 | 80.9 ± 5.6 |
| 850 | 97.8 ± 0.8 | 96.1 ± 1.2 | 88.6 ± 3.1 | 80.7 ± 4.8 |
| 900 | 98.0 ± 0.9 | 96.0 ± 1.4 | 89.2 ± 3.0 | 81.5 ± 5.7 |
| 950 | 97.5 ± 0.8 | 95.7 ± 1.4 | 87.4 ± 4.5 | 80.1 ± 5.9 |
| 1000 | 100.0 ± 0.0 | 99.6 ± 0.4 | 99.7 ± 0.3 | 95.7 ± 3.1 |
| 1050 | 100.0 ± 0.0 | 99.4 ± 0.8 | 99.7 ± 0.5 | 93.8 ± 7.2 |
| 1100 | 100.0 ± 0.0 | 99.1 ± 1.3 | 99.8 ± 0.3 | 92.1 ± 5.9 |
| 1150 | 100.0 ± 0.0 | 99.3 ± 0.5 | 99.6 ± 0.3 | 91.9 ± 4.0 |
| 1200 | 100.0 ± 0.0 | 99.2 ± 0.5 | 99.0 ± 1.0 | 89.8 ± 6.4 |