Embodied Reasoning on ERQA
54.5AccuracyHY-Embodied
Evaluation results
| Method | Links | |
|---|---|---|
| HY-EmbodiedModel Version=0.5 MoT-2B, Thinking Mode=thinking2026.04 | 54.5 | |
| QwenModel Version=3-VL, Number of Parameters=4B2026.04 | 47.3 | |
| MiMo-EmbodiedNumber of Parameters=7B2026.04 | 46.8 | |
| RoboBrainModel Version=2.5, Number of Parameters=4B2026.04 | 43.3 | |
| QwenModel Version=3-VL, Number of Parameters=2B2026.04 | 41.8 | |
| Xiaomi-Robotics-0pre-training=with VL data2026.02 | 40.8 | |
| Qwen3-VL-4B-Instruct2026.02 | 40 | |
| MolmoAct2026.02 | 33.5 | |
| pi_02026.02 | 0 | |
| pi_0.52026.02 | 0 | |
| Xiaomi-Robotics-0pre-training=without VL data2026.02 | 0 |