s0_e4
done · peak unseen_mismatch 89.9
α schedule
α 0.0 ×50 α 0.1739 ×50 α 0.3478 ×50 α 0.5217 ×50 α 0.6957 ×50 α 0.8696 ×50 α 1.0435 ×50 α 1.2174 ×50 α 1.3913 ×50 α 1.5652 ×50 α 1.7391 ×50 α 1.913 ×50 α 2.087 ×50 α 2.2609 ×50 α 2.4348 ×50 α 2.6087 ×50 α 2.7826 ×50 α 2.9565 ×50 α 3.1304 ×50 α 3.3043 ×50 α 3.4783 ×50 α 3.6522 ×50 α 3.8261 ×50 α 4.0 ×50
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 | 50.4 ± 1.6 | 49.9 ± 1.1 | 49.8 ± 1.9 | 49.6 ± 1.9 |
| 100 | 50.0 ± 1.4 | 49.9 ± 1.6 | 49.5 ± 1.0 | 49.5 ± 1.2 |
| 150 | 50.4 ± 2.3 | 49.3 ± 2.7 | 50.9 ± 3.2 | 49.0 ± 1.9 |
| 200 | 60.6 ± 16.7 | 45.5 ± 6.9 | 59.3 ± 17.0 | 45.4 ± 7.0 |
| 250 | 69.2 ± 23.2 | 41.7 ± 15.2 | 68.8 ± 23.2 | 41.3 ± 15.8 |
| 300 | 76.3 ± 24.0 | 38.5 ± 17.6 | 76.3 ± 23.8 | 40.1 ± 18.3 |
| 350 | 81.4 ± 23.2 | 33.6 ± 18.2 | 81.0 ± 23.5 | 33.3 ± 17.3 |
| 400 | 86.9 ± 20.1 | 34.7 ± 18.3 | 86.2 ± 21.2 | 35.9 ± 18.9 |
| 450 | 99.3 ± 2.1 | 26.4 ± 22.8 | 99.3 ± 2.2 | 28.2 ± 23.8 |
| 500 | 100.0 ± 0.0 | 43.0 ± 28.9 | 100.0 ± 0.0 | 43.2 ± 28.9 |
| 550 | 100.0 ± 0.0 | 52.8 ± 30.1 | 100.0 ± 0.0 | 53.9 ± 30.1 |
| 600 | 100.0 ± 0.0 | 63.1 ± 33.4 | 100.0 ± 0.0 | 63.1 ± 32.4 |
| 650 | 100.0 ± 0.0 | 71.7 ± 32.7 | 100.0 ± 0.1 | 71.0 ± 31.3 |
| 700 | 99.7 ± 0.8 | 88.2 ± 22.0 | 98.8 ± 2.8 | 86.5 ± 21.0 |
| 750 | 99.6 ± 0.7 | 91.5 ± 16.1 | 98.5 ± 2.3 | 89.4 ± 16.2 |
| 800 | 98.8 ± 1.1 | 93.2 ± 12.1 | 96.6 ± 3.3 | 89.9 ± 11.3 |
| 850 | 97.3 ± 1.6 | 92.3 ± 12.0 | 93.9 ± 4.4 | 87.8 ± 11.1 |
| 900 | 96.6 ± 1.3 | 95.3 ± 1.4 | 91.5 ± 4.3 | 89.0 ± 3.6 |
| 950 | 95.0 ± 1.6 | 93.9 ± 1.5 | 87.4 ± 5.9 | 84.7 ± 5.1 |
| 1000 | 93.4 ± 1.2 | 92.0 ± 1.5 | 83.6 ± 4.1 | 80.7 ± 4.4 |
| 1050 | 92.4 ± 1.5 | 91.2 ± 1.5 | 80.9 ± 4.6 | 77.9 ± 3.8 |
| 1100 | 91.3 ± 1.1 | 90.0 ± 1.3 | 78.2 ± 4.2 | 75.2 ± 2.9 |
| 1150 | 89.8 ± 1.1 | 89.1 ± 1.0 | 74.8 ± 4.6 | 72.2 ± 3.9 |
| 1200 | 89.2 ± 1.1 | 88.0 ± 1.2 | 73.5 ± 4.1 | 70.9 ± 3.7 |