Multimodal Agent Performance on HR-Bench 8k
78.2AccuracyBaseline
Evaluation Results
| Method | Links | |
|---|---|---|
| BaselineBackbone=235B-FP8, Training regime=None2026.06 | 78.2 | |
| ToolGateBackbone=235B-FP8, Training regime=Cross-domain2026.06 | 77 | |
| ToolGateBackbone=30B, Training regime=In-domain2026.06 | 75.5 | |
| ToolGateBackbone=30B, Training regime=Cross-domain2026.06 | 74.3 | |
| BaselineBackbone=30B, Training regime=None2026.06 | 72.6 |