Reasoning Efficiency on ScienceWorld (Seen)
83.16Success Rate (%)BPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BPOSize=8B2025.08 | 83.16 | 112 | |
| Qwen-3-ThinkingSize=8B2025.08 | 57.45 | 763 | |
| Deepseek-R1Size=671B2025.08 | 56.7 | 620 |
| Method | Links | ||
|---|---|---|---|
| BPOSize=8B2025.08 | 83.16 | 112 | |
| Qwen-3-ThinkingSize=8B2025.08 | 57.45 | 763 | |
| Deepseek-R1Size=671B2025.08 | 56.7 | 620 |