s1_e2_blk
done · peak unseen_mismatch 98.4
α schedule
α 1.0 ×600 α 2.0 ×600
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 2.0
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 | 50.6 ± 1.3 | 49.1 ± 0.7 | 50.0 ± 2.0 | 50.2 ± 1.2 |
| 100 | 50.1 ± 1.7 | 50.3 ± 2.2 | 50.3 ± 1.8 | 50.0 ± 1.8 |
| 150 | 51.2 ± 1.8 | 49.7 ± 1.9 | 51.1 ± 1.2 | 48.5 ± 1.1 |
| 200 | 57.8 ± 15.6 | 46.8 ± 8.9 | 58.3 ± 15.1 | 47.6 ± 9.7 |
| 250 | 71.9 ± 20.1 | 41.9 ± 10.9 | 72.1 ± 19.3 | 41.2 ± 9.8 |
| 300 | 88.1 ± 18.5 | 32.0 ± 21.6 | 88.1 ± 17.7 | 32.5 ± 20.6 |
| 350 | 100.0 ± 0.0 | 39.0 ± 27.4 | 100.0 ± 0.0 | 39.3 ± 27.7 |
| 400 | 100.0 ± 0.0 | 36.4 ± 24.8 | 100.0 ± 0.0 | 36.3 ± 25.2 |
| 450 | 100.0 ± 0.0 | 32.9 ± 22.1 | 100.0 ± 0.0 | 33.5 ± 22.8 |
| 500 | 100.0 ± 0.0 | 45.8 ± 30.8 | 100.0 ± 0.0 | 46.8 ± 30.7 |
| 550 | 100.0 ± 0.0 | 47.6 ± 22.1 | 100.0 ± 0.0 | 48.1 ± 22.2 |
| 600 | 100.0 ± 0.0 | 52.0 ± 27.8 | 100.0 ± 0.0 | 53.1 ± 27.6 |
| 650 | 100.0 ± 0.0 | 79.2 ± 22.7 | 100.0 ± 0.0 | 79.8 ± 22.0 |
| 700 | 100.0 ± 0.1 | 74.0 ± 27.7 | 99.7 ± 0.9 | 73.8 ± 26.9 |
| 750 | 99.9 ± 0.2 | 90.1 ± 14.7 | 99.6 ± 1.2 | 89.8 ± 13.9 |
| 800 | 99.8 ± 0.5 | 95.9 ± 5.8 | 99.4 ± 1.7 | 95.8 ± 5.1 |
| 850 | 99.6 ± 0.7 | 99.1 ± 1.1 | 99.4 ± 1.3 | 98.4 ± 2.0 |
| 900 | 99.0 ± 1.0 | 98.2 ± 1.8 | 97.8 ± 2.3 | 96.3 ± 3.6 |
| 950 | 98.7 ± 1.3 | 97.9 ± 1.7 | 96.5 ± 3.1 | 94.5 ± 4.6 |
| 1000 | 98.4 ± 1.2 | 97.3 ± 1.7 | 94.9 ± 3.8 | 92.7 ± 4.2 |
| 1050 | 98.0 ± 1.1 | 97.1 ± 1.3 | 94.4 ± 2.7 | 91.8 ± 3.7 |
| 1100 | 97.8 ± 1.0 | 96.6 ± 1.4 | 93.8 ± 3.3 | 90.8 ± 4.1 |
| 1150 | 97.5 ± 1.1 | 96.3 ± 1.4 | 92.8 ± 4.4 | 90.0 ± 4.9 |
| 1200 | 97.3 ± 1.3 | 96.2 ± 1.4 | 90.8 ± 3.8 | 89.1 ± 3.7 |