Embodied Question Answering on BridgeEQA 1,100 QA pairs (test)
64.8Answer CorrectnessEMVR VLM w/ Images + SG
Evaluation Results
| Method | Links | |
|---|---|---|
| EMVR VLM w/ Images + SGVision-Language Model=Grok 4 Fast2025.11 | 64.8 | |
| EMVR VLM w/ SG OnlyVision-Language Model=Grok 4 Fast2025.11 | 63.8 | |
| Multi-Frame VLM w/ SGVision-Language Model=Grok 4 Fast2025.11 | 62.2 | |
| Socratic LLM w/ SGVision-Language Model=Grok 4 Fast2025.11 | 61.4 | |
| EMVR VLM w/ SG OnlyVision-Language Model=Gemini 2.5 Flash2025.11 | 60.9 | |
| Socratic LLM w/ SGVision-Language Model=Gemini 2.5 Flash2025.11 | 58.8 | |
| Multi-Frame VLM w/ SGVision-Language Model=Gemini 2.5 Flash Lite2025.11 | 58.1 | |
| Multi-Frame VLMVision-Language Model=Grok 4 Fast2025.11 | 57.6 | |
| EMVR VLM w/ Images + SGVision-Language Model=Gemini 2.5 Flash2025.11 | 55.1 | |
| Multi-Frame VLM w/ SGVision-Language Model=Gemini 2.5 Flash2025.11 | 54.8 | |
| Socratic LLM w/ SGVision-Language Model=Gemini 2.5 Flash Lite2025.11 | 54.2 | |
| EMVR VLM w/ SG OnlyVision-Language Model=Gemini 2.5 Flash Lite2025.11 | 51.2 | |
| Multi-Frame VLMVision-Language Model=Gemini 2.5 Flash Lite2025.11 | 50.7 | |
| EMVR VLM w/ Images + SGVision-Language Model=Gemini 2.5 Flash Lite2025.11 | 49.7 | |
| Multi-Frame VLMVision-Language Model=Gemini 2.5 Flash2025.11 | 48.4 |