teeth_6
done · peak unseen_mismatch 98.3
α schedule
α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 3.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 2.6571 ×19 α 2.6571 ×19 α 2.6571 ×19 α 2.6571 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 2.3143 ×19 α 2.3143 ×19 α 2.3143 ×19 α 2.3143 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.9714 ×19 α 1.9714 ×19 α 1.9714 ×19 α 1.9714 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.0 ×19 α 1.6286 ×19 α 1.6286 ×19 α 1.6286 ×19 α 1.6286 ×19 α 1.0 ×18 α 1.0 ×18 α 1.0 ×18 α 1.0 ×18 α 1.2857 ×18 α 1.2857 ×18 α 1.2857 ×18 α 1.2857 ×18 α 1.0 ×18 α 1.0 ×18 α 1.0 ×18 α 1.0 ×18
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 |
|---|---|---|---|---|
| 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 | 51.5 ± 1.8 | 50.5 ± 1.9 | 50.8 ± 1.9 | 49.5 ± 1.3 |
| 125 | 56.3 ± 7.1 | 50.3 ± 3.5 | 54.8 ± 7.1 | 49.7 ± 2.5 |
| 150 | 71.9 ± 19.1 | 52.3 ± 10.9 | 70.2 ± 19.8 | 52.0 ± 9.9 |
| 175 | 87.0 ± 17.6 | 61.6 ± 19.0 | 86.0 ± 18.3 | 60.6 ± 18.4 |
| 200 | 94.0 ± 13.1 | 69.8 ± 26.0 | 93.3 ± 13.4 | 68.1 ± 25.7 |
| 225 | 96.6 ± 11.4 | 81.7 ± 18.2 | 96.0 ± 10.8 | 79.8 ± 17.9 |
| 250 | 97.5 ± 10.7 | 85.3 ± 18.0 | 97.5 ± 10.3 | 83.8 ± 18.3 |
| 275 | 97.6 ± 10.4 | 82.5 ± 22.5 | 97.4 ± 10.7 | 81.3 ± 22.3 |
| 300 | 98.6 ± 6.0 | 80.4 ± 20.0 | 98.5 ± 6.3 | 78.5 ± 20.8 |
| 325 | 99.9 ± 0.3 | 90.6 ± 18.3 | 99.3 ± 1.8 | 88.8 ± 18.9 |
| 350 | 99.5 ± 0.8 | 89.2 ± 20.6 | 98.3 ± 3.0 | 86.4 ± 19.9 |
| 375 | 99.3 ± 0.9 | 96.2 ± 5.1 | 97.8 ± 2.6 | 93.0 ± 6.9 |
| 400 | 100.0 ± 0.0 | 98.4 ± 3.9 | 99.9 ± 0.1 | 97.7 ± 4.4 |
| 450 | 100.0 ± 0.0 | 98.2 ± 5.3 | 99.8 ± 0.6 | 97.1 ± 5.7 |
| 500 | 99.1 ± 0.8 | 97.7 ± 1.1 | 96.2 ± 3.7 | 93.2 ± 4.9 |
| 550 | 99.9 ± 0.1 | 99.5 ± 0.9 | 99.7 ± 0.3 | 98.3 ± 2.0 |
| 600 | 100.0 ± 0.1 | 99.6 ± 0.7 | 99.5 ± 0.5 | 97.4 ± 2.8 |
| 650 | 99.0 ± 0.5 | 97.6 ± 1.6 | 96.3 ± 2.7 | 91.6 ± 6.4 |
| 700 | 99.9 ± 0.1 | 99.6 ± 0.5 | 99.5 ± 0.7 | 96.7 ± 4.2 |
| 750 | 100.0 ± 0.0 | 99.2 ± 1.3 | 99.6 ± 0.6 | 95.2 ± 5.3 |
| 800 | 99.7 ± 0.2 | 98.7 ± 1.2 | 98.2 ± 1.4 | 92.2 ± 6.4 |
| 850 | 100.0 ± 0.0 | 99.6 ± 0.7 | 99.7 ± 0.5 | 96.0 ± 4.4 |
| 900 | 100.0 ± 0.1 | 98.9 ± 1.7 | 99.1 ± 1.1 | 93.4 ± 6.7 |