Path-level reasoning on UNOBench synthetic Medium (test)
74.8SR-P (%)UNOGrasp
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| UNOGraspPrompting Type=Oracle (with SoM)2025.11 | 74.8 | 67.2 | 69.1 | 0.37 | |
| UNOGraspPrompting Type=Natural Language Prompting2025.11 | 68 | 55.8 | 59.7 | 0.53 | |
| Qwen2.5-VLPrompting Type=Oracle (with SoM), Evaluation Protocol=SFT2025.11 | 64.2 | 53.1 | 56.5 | — | |
| Qwen2.5-VLPrompting Type=Natural Language Prompting, Evaluation Protocol=SFT2025.11 | 59.6 | 47.9 | 51.5 | — | |
| Gemini Robotics-ER 1.5Prompting Type=Natural Language Prompting, Evaluation Protocol=ICL2025.11 | 45.5 | 37.3 | 39.5 | 0.85 | |
| Gemini Robotics-ER 1.5Prompting Type=Oracle (with SoM), Evaluation Protocol=ICL2025.11 | 41.8 | 39.7 | 38.7 | 0.5 | |
| Gemini Robotics-ER 1.5Prompting Type=Natural Language Prompting2025.11 | 36.9 | 30.9 | 32.5 | 0.87 | |
| Gemini Robotics-ER 1.5Prompting Type=Oracle (with SoM)2025.11 | 33 | 30.2 | 29.8 | 0.56 | |
| Qwen2.5-VLPrompting Type=Oracle (with SoM), Evaluation Protocol=ICL2025.11 | 20.7 | 27.2 | 20.3 | 0.73 | |
| Qwen2.5-VLPrompting Type=Natural Language Prompting, Evaluation Protocol=ICL2025.11 | 11.8 | 10.8 | 10.6 | 0.88 |