Share your thoughts, 1 month free Claude Pro on usSee more

Goal reasoning on EgoToM

89Accuracy

Humans

Updated 2mo ago

Evaluation Results

Method	Links
Humans 2026.03		89
Qwen2.5-VL-7B 2026.03		88.9
Qwen2.5-VL-7B 2026.03		88.9
Qwen2.5-VL-7B 2026.03		88.4
Humans 2026.03		88
Qwen2.5-VL-7B 2026.03		87.1
Qwen2.5-VL-7B 2026.03		86.9
Gemini-2.5-Flash 2026.03		86
Video-Llama2-72B 2026.03		85
GPT-4-Turbo 2026.03		83
Humans 2026.03		78
Humans 2026.03		77
LLaVA-Next-Video-7B 2026.03		74.5
LLaVA-Next-Video-7B 2026.03		73.2
CogVLM2 2026.03		73
LLaVA-Next-Video-7B 2026.03		72.6
GPT-4o 2026.03		68.7
LLaVA-Next-Video-7B 2026.03		62.1
LLaVA-Next-Video-7B 2026.03		61.5
Qwen2.5-VL-7B 2026.03		60.3
LLaVA-Next-Video-7B 2026.03		50.6
GPT-4-Turbo 2026.03		42
Gemini-2.5-Flash 2026.03		40.1
Video-Llama2-72B 2026.03		40
Qwen2.5-VL-7B 2026.03		37.6
Qwen2.5-VL-7B 2026.03		37.5
CogVLM2 2026.03		36
Qwen2.5-VL-7B 2026.03		33.2
Qwen2.5-VL-7B 2026.03		31.9
Qwen2.5-VL-7B 2026.03		31.1
LLaVA-Next-Video-7B 2026.03		29.7
LLaVA-Next-Video-7B 2026.03		29
LLaVA-Next-Video-7B 2026.03		25.4
LLaVA-Next-Video-7B 2026.03		25.3
LLaVA-Next-Video-7B 2026.03		24
GPT-4o 2026.03		22.6
Qwen2.5-VL-7B 2026.03		14.3
LLaVA-Next-Video-7B 2026.03		10.1