Multi-Hop Question Answering on MuSiQue (Score)
37.8ScoreToolForge
Evaluation Results
| Method | Links | |
|---|---|---|
| ToolForgeBase Model=Qwen3-8B2026.04 | 37.8 | |
| Dr.ZeroBase Model=Qwen3-4B-Instruct-25072026.04 | 14.4 | |
| π-PlayBase Model=Qwen3-4B-Instruct-25072026.04 | 13.4 | |
| Dr.ZeroBase Model=Qwen3-8B2026.04 | 13.1 | |
| Search-R1Base Model=Qwen3-4B-Instruct-25072026.04 | 12.9 | |
| π-PlayBase Model=Qwen3-8B2026.04 | 12.4 | |
| π-PlayBase Model=Qwen3-4B2026.04 | 11.2 | |
| SQLM*Base Model=Qwen3-4B2026.04 | 10.8 | |
| Search-R1Base Model=Qwen3-4B2026.04 | 10.5 | |
| SQLM*Base Model=Qwen3-4B-Instruct-25072026.04 | 10.2 | |
| SQLM*Base Model=Qwen3-8B2026.04 | 9.9 | |
| Search-R1Base Model=Qwen3-8B2026.04 | 9.8 | |
| Dr.ZeroBase Model=Qwen3-4B2026.04 | 7.8 | |
| ReActBase Model=Qwen3-8B2026.04 | 7.4 | |
| ReActBase Model=Qwen3-4B-Instruct-25072026.04 | 6.3 | |
| ReActBase Model=Qwen3-4B2026.04 | 1.6 |