Object-level reasoning on UNOBench Easy synthetic (test)
81.3OPUNOGrasp
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| UNOGraspPrompting Strategy=Oracle (with SoM)2025.11 | 81.3 | 85.9 | 82.6 | |
| Gemini Robotics-ER 1.5Prompting Strategy=Oracle (with SoM), Evaluation Mode=In-Context Learning (ICL)2025.11 | 67.8 | 76.5 | 70.4 | |
| UNOGraspPrompting Strategy=Natural Language Prompting2025.11 | 65.6 | 67.4 | 66.1 | |
| Gemini Robotics-ER 1.5Prompting Strategy=Oracle (with SoM), Evaluation Mode=Base Prompting2025.11 | 56 | 62.3 | 57.9 | |
| Qwen2.5-VLPrompting Strategy=Oracle (with SoM), Evaluation Mode=In-Context Learning (ICL)2025.11 | 19.6 | 33.3 | 22.8 | |
| Gemini Robotics-ER 1.5Prompting Strategy=Natural Language Prompting, Evaluation Mode=In-Context Learning (ICL)2025.11 | 3.4 | 3.7 | 3.5 | |
| Gemini Robotics-ER 1.5Prompting Strategy=Natural Language Prompting, Evaluation Mode=Base Prompting2025.11 | 2.5 | 2.7 | 2.6 | |
| Qwen2.5-VLPrompting Strategy=Natural Language Prompting, Evaluation Mode=In-Context Learning (ICL)2025.11 | 2.1 | 3.6 | 2.5 |