Representation Learning on dSprites
73.9k-NN AccuracyBLAE
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| BLAE2026.04 | 73.9 | 0.0064 | 0.004 | 0 | 0.0169 | |
| SPAE2026.04 | 72.4 | 0.0295 | 0.0091 | 0.0001 | 0.0172 | |
| CAE2026.04 | 67.9 | 0.0399 | 0.0205 | 0.0018 | 0.0259 | |
| TAE2026.04 | 55.8 | 0.0315 | 0.0146 | 0.0013 | 0.0284 | |
| GRAE2026.04 | 52.3 | 0.0294 | 0.0091 | 0.0001 | 0.0306 | |
| GAE2026.04 | 49.9 | 0.0373 | 0.0179 | 0.0017 | 0.0436 | |
| IRAE2026.04 | 48.9 | 0.111 | 0.0313 | 0.0033 | 0.0557 | |
| Vanilla AE2026.04 | 48.9 | 0.121 | 0.0523 | 0.0047 | 0.0479 | |
| DN2026.04 | 48.1 | 0.0841 | 0.0732 | 0.0064 | 0.0347 | |
| GGAE2026.04 | 43.6 | 0.183 | 0.295 | 0.0027 | 0.0519 |