Goal-conditioned Generalization on GSR-bench Medium 1.0 (test)
29.1Task ProgressRoBoSR-8B-FT
Evaluation Results
| Method | Links | |
|---|---|---|
| RoBoSR-8B-FTmode=fine-tuned, demonstration_trajectories=362026.06 | 29.1 | |
| RoBoSR-8Bmode=zero-shot2026.06 | 19 | |
| DeepSeek-V3type=prompting-based2026.06 | 9 | |
| Qwen3-8Btype=prompting-based2026.06 | 6.7 | |
| Claude-sonnet-4.5type=prompting-based2026.06 | 6 | |
| Gemini-2.5-protype=prompting-based2026.06 | 5 | |
| GPT-5type=prompting-based2026.06 | 5 |