Vision-Language Reasoning on NaturalBench (test)
66.02Simple Accuracy
Q4 system redistr (prop)
Updated 4d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.01 | 66.02 | 34.59 | 66.67 | |
2026.01 | 62.17 | 25.52 | 82.07 | |
2026.01 | 61.28 | 23.72 | 83.83 | |
2026.01 | 60.97 | 23 | 84.55 | |
2026.01 | 60.22 | 21.55 | 85.67 | |
2026.01 | 59.97 | 21.1 | 86.03 | |
2026.01 | 59.79 | 20.66 | 86.34 |