s3_e4_blk
done · peak unseen_mismatch 76.2
α schedule
α 3.0 ×600 α 4.0 ×600
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 4.0
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 | 51.3 ± 2.2 | 49.6 ± 0.8 | 50.0 ± 1.7 | 50.3 ± 1.6 |
| 100 | 55.9 ± 7.4 | 53.6 ± 4.6 | 52.7 ± 5.7 | 50.5 ± 3.7 |
| 150 | 76.3 ± 12.5 | 72.0 ± 11.7 | 69.1 ± 13.0 | 63.1 ± 11.1 |
| 200 | 88.2 ± 4.3 | 86.4 ± 4.6 | 78.2 ± 8.2 | 75.1 ± 6.9 |
| 250 | 90.7 ± 1.6 | 89.4 ± 1.6 | 79.0 ± 4.9 | 76.2 ± 3.6 |
| 300 | 91.8 ± 1.5 | 90.3 ± 1.5 | 78.3 ± 4.3 | 75.5 ± 4.1 |
| 350 | 91.7 ± 1.0 | 90.3 ± 1.2 | 77.0 ± 3.1 | 73.9 ± 2.3 |
| 400 | 92.6 ± 1.2 | 90.7 ± 1.6 | 79.5 ± 5.0 | 74.6 ± 4.8 |
| 450 | 92.4 ± 0.8 | 90.8 ± 1.4 | 77.1 ± 3.2 | 73.7 ± 2.8 |
| 500 | 93.2 ± 0.9 | 91.1 ± 1.3 | 80.0 ± 3.3 | 73.9 ± 4.2 |
| 550 | 92.5 ± 1.2 | 90.8 ± 1.4 | 78.1 ± 4.4 | 72.6 ± 5.7 |
| 600 | 93.4 ± 0.8 | 91.5 ± 1.3 | 79.0 ± 3.7 | 73.8 ± 4.2 |
| 650 | 91.1 ± 1.5 | 89.7 ± 1.3 | 76.6 ± 4.2 | 72.3 ± 4.7 |
| 700 | 91.3 ± 1.4 | 89.3 ± 1.6 | 76.7 ± 3.5 | 71.9 ± 3.8 |
| 750 | 90.8 ± 1.7 | 89.6 ± 1.9 | 76.3 ± 5.1 | 72.2 ± 5.3 |
| 800 | 91.0 ± 1.6 | 89.8 ± 0.9 | 76.6 ± 3.1 | 72.6 ± 3.2 |
| 850 | 91.2 ± 1.1 | 89.5 ± 1.2 | 76.8 ± 2.8 | 73.0 ± 3.8 |
| 900 | 90.9 ± 0.8 | 89.3 ± 1.3 | 74.2 ± 4.0 | 70.9 ± 5.8 |
| 950 | 90.7 ± 1.2 | 89.6 ± 1.4 | 75.4 ± 4.0 | 70.5 ± 4.4 |
| 1000 | 91.2 ± 1.5 | 90.1 ± 1.8 | 76.8 ± 4.0 | 72.4 ± 5.7 |
| 1050 | 91.7 ± 1.3 | 90.7 ± 1.3 | 76.6 ± 3.4 | 73.0 ± 3.9 |
| 1100 | 91.4 ± 1.1 | 90.3 ± 1.3 | 77.4 ± 3.3 | 72.9 ± 4.7 |
| 1150 | 91.0 ± 1.0 | 90.0 ± 1.5 | 76.7 ± 2.7 | 72.7 ± 5.1 |
| 1200 | 91.1 ± 0.4 | 89.5 ± 1.6 | 75.2 ± 2.7 | 71.4 ± 5.7 |