Visual Reasoning on ARC 1.0 (test)
98AccuracyBest human
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Best human2026.01 | 98 | — | — | |
| c-d-4-TTT3-Unet-ensem#param=83M2026.01 | 62.6 | — | — | |
| VARC-ViT-Unet-ensem#param=73M2026.01 | 60.4 | — | — | |
| Average human2026.01 | 60.2 | — | — | |
| c-d-4-TTT3#param=28M2026.01 | 58.8 | — | — | |
| VARC-ViT#param=18M2026.01 | 54.5 | — | — | |
| VARC-ViT-large#param=66M2026.01 | 53 | — | — | |
| TRM#param=7M2026.01 | 44.6 | — | — | |
| HRM#param=27M2026.01 | 40.3 | — | — | |
| HRMParam=27M2026.06 | — | — | 41.25 | |
| L-VARCParam=18M2026.06 | — | 50.62 | 55.37 | |
| TRMParam=7M2026.06 | — | — | 43.5 | |
| VARCParam=18M2026.06 | — | 49.75 | 55.63 |