Theory of Mind reasoning on MMToM-QA Multimodal benchmark leaderboard · SOTA2 Research

Theory of Mind reasoning on MMToM-QA Multimodal

95.8Belief Inference 1.1Human

Evaluation Results

Method
Human2024.01	95.8	96.7	100	97.5	90	91.7	83.3	88.9	88.5	93
Human2025.06	95.8	96.7	100	97.5	90	91.7	83.3	88.9	88.5	93
GPT-4V2024.01	94	13	59	55.3	56	26.7	4	52	34.7	44
GPT-4V2025.06	94	13	59	55.3	56	26.7	4	52	34.7	44
Scalable Bayesian ToM Planner2025.06	92.1	76	93	87.1	73.4	80	75.5	78.7	76.9	81.3
BIP-ALM2024.01	90	69	86	81.7	68	78.7	56	73.3	69	75.3
BIPALM w/ GPT-J-6B2025.06	90	69	86	81.7	68	78.7	56	73.3	69	75.3
BIP-ALM2024.01	88	68	85	80.3	62.7	77.3	72	80	73.3	76.7
BIPALM w/ Llama2-7B2025.06	88	68	85	80.3	62.7	77.3	72	80	73.3	76.7
InstructBLIP2024.01	62	52	32	48.7	46.7	29.3	42.7	60	44.7	46.7
LLaVA2024.01	46	14	69	43	65.3	22.7	40	48	44	43.5
LLaVA-7B2025.06	46	14	69	43	65.3	22.7	40	48	44	43.5
Video-LLaMA 22024.01	36	38	52	42	36	41.3	30.7	45.3	38.3	40.2
Video-Llama2-13B2025.06	36	38	52	42	36	41.3	30.7	45.3	38.3	40.2