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

Theory of Mind reasoning on MMToM-QA Video-only

69.1Belief Inference 1.1

Human

Updated 22d ago

Evaluation Results

Method	Links
Human 2024.01		69.1	64.3	86.4	73.3	58.5	60	76.7	63.3	64.6	68.9
Human 2025.06		69.1	64.3	86.4	73.3	58.5	60	76.7	63.3	64.6	68.9
BIP-ALM 2024.01		69	63	60	64	62.7	54.7	53.3	62.7	58.3	61.2
BIPALM w/ Llama2-7B 2025.06		69	63	60	64	62.7	54.7	53.3	62.7	58.3	61.2
GPT-4V 2024.01		64	34	39	45.7	54.7	26.7	48	56	46.3	46
GPT-4V 2025.06		64	34	39	45.7	54.7	26.7	48	56	46.3	46
BIP-ALM 2024.01		63	57	72	64	45.3	62.7	50.7	62.7	55.3	59.7
BIPALM w/ GPT-J-6B 2025.06		63	57	72	64	45.3	62.7	50.7	62.7	55.3	59.7
InstructBLIP 2024.01		56	50	42	49.3	56	45.3	54.7	53.3	52.3	50.8
LLaVA 2024.01		33	15	69	39	44	24	56	57.3	45.3	42.2
LLaVA-7B 2025.06		33	15	69	39	44	24	56	57.3	45.3	42.2
Video-LLaMA 2 2024.01		24	32	67	41	50.7	45.3	56	52	51	46
Video-Llama2-13B 2025.06		24	32	67	41	50.7	45.3	56	52	51	46