v_0.0
done · peak unseen_mismatch 94.2
Referenced by
α schedule
α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 2.1429 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 0.0 ×19 α 1.2857 ×19 α 1.2857 ×19 α 1.2857 ×19 α 1.2857 ×19 α 1.2857 ×18 α 1.2857 ×18 α 1.2857 ×18 α 1.2857 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18 α 0.0 ×18
Figures
Results
accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 0.0
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 |
|---|---|---|---|---|
| 25 | 51.4 ± 1.5 | 49.5 ± 1.6 | 49.8 ± 1.9 | 49.7 ± 2.1 |
| 50 | 50.7 ± 1.9 | 49.6 ± 1.4 | 50.1 ± 1.5 | 50.8 ± 1.1 |
| 75 | 51.9 ± 1.8 | 50.1 ± 1.9 | 50.5 ± 1.7 | 50.0 ± 1.9 |
| 100 | 53.8 ± 2.9 | 52.1 ± 2.9 | 51.5 ± 1.8 | 50.0 ± 1.5 |
| 125 | 65.8 ± 10.5 | 60.5 ± 8.1 | 59.9 ± 9.2 | 54.9 ± 5.3 |
| 150 | 79.3 ± 12.4 | 73.6 ± 12.8 | 72.2 ± 14.1 | 66.3 ± 12.7 |
| 175 | 92.0 ± 12.8 | 73.8 ± 19.5 | 90.1 ± 14.0 | 71.5 ± 19.1 |
| 200 | 97.1 ± 10.9 | 62.1 ± 30.1 | 97.3 ± 10.5 | 61.5 ± 29.8 |
| 225 | 97.5 ± 11.1 | 54.5 ± 34.5 | 97.5 ± 10.9 | 54.0 ± 34.2 |
| 250 | 97.5 ± 10.8 | 52.2 ± 32.6 | 97.6 ± 10.4 | 51.9 ± 33.0 |
| 275 | 97.5 ± 10.8 | 40.2 ± 34.7 | 97.5 ± 10.7 | 40.9 ± 34.6 |
| 300 | 97.8 ± 9.4 | 42.2 ± 32.4 | 97.8 ± 9.5 | 42.1 ± 32.2 |
| 325 | 100.0 ± 0.1 | 50.4 ± 37.2 | 100.0 ± 0.1 | 50.8 ± 36.8 |
| 350 | 100.0 ± 0.0 | 75.7 ± 28.7 | 100.0 ± 0.0 | 75.6 ± 28.3 |
| 375 | 99.9 ± 0.1 | 87.8 ± 15.5 | 99.8 ± 0.6 | 86.8 ± 15.5 |
| 400 | 99.6 ± 0.5 | 94.3 ± 8.6 | 98.8 ± 1.9 | 92.1 ± 9.2 |
| 450 | 99.2 ± 0.9 | 95.3 ± 8.1 | 97.5 ± 2.8 | 91.0 ± 9.1 |
| 500 | 100.0 ± 0.0 | 94.1 ± 13.7 | 100.0 ± 0.0 | 94.2 ± 13.4 |
| 550 | 100.0 ± 0.0 | 76.4 ± 27.6 | 100.0 ± 0.0 | 75.3 ± 27.8 |
| 600 | 100.0 ± 0.0 | 69.4 ± 31.8 | 100.0 ± 0.0 | 68.8 ± 31.5 |
| 650 | 100.0 ± 0.0 | 87.8 ± 26.0 | 99.8 ± 0.7 | 86.1 ± 25.9 |
| 700 | 100.0 ± 0.1 | 92.2 ± 20.9 | 99.6 ± 0.6 | 89.2 ± 21.8 |
| 750 | 99.9 ± 0.2 | 96.8 ± 5.4 | 98.9 ± 1.2 | 92.3 ± 8.0 |
| 800 | 100.0 ± 0.0 | 75.4 ± 31.6 | 100.0 ± 0.0 | 74.2 ± 32.7 |
| 850 | 100.0 ± 0.0 | 61.4 ± 34.2 | 100.0 ± 0.0 | 60.5 ± 34.7 |
| 900 | 100.0 ± 0.0 | 59.8 ± 33.1 | 100.0 ± 0.0 | 59.5 ± 32.8 |