Path-level reasoning on UNOBench real (Easy)
76.2SR (Precision)UNOGrasp
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| UNOGraspPrompting=Oracle (with SoM), Protocol=Standard2025.11 | 76.2 | 79 | 77.2 | 0.15 | |
| UNOGraspPrompting=Natural Language Prompting, Protocol=Standard2025.11 | 71.1 | 71.8 | 71.3 | 0.26 | |
| Qwen2.5-VLPrompting=Oracle (with SoM), Protocol=SFT2025.11 | 70.2 | 70.6 | 70.4 | — | |
| Qwen2.5-VLPrompting=Natural Language Prompting, Protocol=SFT2025.11 | 64 | 64.5 | 64.2 | — | |
| Gemini Robotics-ER 1.5Prompting=Natural Language Prompting, Protocol=ICL2025.11 | 60.5 | 63.3 | 61.4 | 0.8 | |
| Gemini Robotics-ER 1.5Prompting=Oracle (with SoM), Protocol=ICL2025.11 | 60.1 | 62.9 | 60.9 | 0.8 | |
| Gemini Robotics-ER 1.5Prompting=Natural Language Prompting, Protocol=Standard2025.11 | 50.3 | 51 | 50.5 | 0.79 | |
| Gemini Robotics-ER 1.5Prompting=Oracle (with SoM), Protocol=Standard2025.11 | 47.7 | 48.3 | 47.9 | 0.8 | |
| Qwen2.5-VLPrompting=Oracle (with SoM), Protocol=ICL2025.11 | 42.4 | 47.2 | 43.5 | 0.37 | |
| Qwen2.5-VLPrompting=Natural Language Prompting, Protocol=ICL2025.11 | 21.3 | 22.6 | 21.7 | 0.79 |