Agentic Search on HotpotQA
0.603String-F1Vanilla-Qwen3-32B
Evaluation Results
| Method | Links | |
|---|---|---|
| Vanilla-Qwen3-32BASP=false2026.04 | 0.603 | |
| Vanilla-Qwen3-8BASP=false2026.04 | 0.582 | |
| Mixed-Qwen3-1.7BASP=true2026.04 | 0.582 | |
| SFT-Qwen3-1.7BASP=true2026.04 | 0.576 | |
| OPD-Qwen3-1.7BASP=true2026.04 | 0.562 | |
| Vanilla-Qwen3-4BASP=false2026.04 | 0.534 | |
| SFT-Llama-3.2-3BASP=true2026.04 | 0.53 | |
| Distilled-Qwen3-1.7BASP=false2026.04 | 0.479 | |
| Distilled-Llama3.2-3BASP=false2026.04 | 0.471 | |
| SFT-Qwen3-0.6BASP=true2026.04 | 0.47 | |
| SFT-Llama-3.2-1BASP=true2026.04 | 0.445 | |
| Vanilla-Qwen3-1.7BASP=false2026.04 | 0.423 | |
| Distilled-Llama3.2-1BASP=false2026.04 | 0.382 | |
| Vanilla-Qwen3-0.6BASP=false2026.04 | 0.194 |