In-the-wild model generalization on Human Bench Text-based Demo
23.4NSE
Qwen3VL-32B
Updated 1mo ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.01 | 23.4 | 77.9 | 0.1 | |
2026.01 | 23.8 | 75.3 | 1.9 | |
2026.01 | 25 | 69.3 | 0.1 | |
2026.01 | 25.7 | 68.5 | 0 | |
2026.01 | 25.9 | 59.7 | 7.7 | |
2026.01 | 26.7 | 70.5 | 0.2 | |
2026.01 | 29.3 | 58.7 | 11.4 | |
2026.01 | 29.5 | 55.7 | 2.6 | |
2026.01 | 30 | 53.8 | 0.6 | |
2026.01 | 30.9 | 46 | 11.3 | |
2026.01 | 32.4 | 49.6 | 5.8 | |
2026.01 | 33.6 | 35.3 | 0.9 | |
2026.01 | 52.7 | 17.2 | 23.8 | |
2026.01 | 67.4 | 5.6 | 0 |