Search-augmented Multi-turn Tool Calling on MuSiQue (Success Rate)
13.73Success RateENVRL-GiGPO
Evaluation Results
| Method | Links | |
|---|---|---|
| ENVRL-GiGPOBackbone=Qwen2.5-3B-Instruct2026.06 | 13.73 | |
| GiGPOBackbone=Qwen2.5-3B-Instruct2026.06 | 13.07 |
| Method | Links | |
|---|---|---|
| ENVRL-GiGPOBackbone=Qwen2.5-3B-Instruct2026.06 | 13.73 | |
| GiGPOBackbone=Qwen2.5-3B-Instruct2026.06 | 13.07 |