Robot-Centric MindPower Reasoning on MindPower Video-input (test) benchmark leaderboard · SOTA2 Research

Robot-Centric MindPower Reasoning on MindPower Video-input (test)

49.14Belief (BERTScore)Ours (SFT+Mind-Reward)

Evaluation Results

Method	Links
Ours (SFT+Mind-Reward)2025.11		49.14	44.79	59.93	46.49	51.25	45.75	37.79	42.57	40.17	47.12	11.75	15.4	8.87
Human Baseline2025.11		47.65	—	—	61.81	46.76	53.71	39.18	52.93	34.55	56.66	19.37	26.26	8.19
SFT only2025.11		43.15	32.78	52.72	42.48	47.01	37.83	34.86	39.48	36.7	43.84	8.5	10.48	8.78
Gemini-2.5 Pro2025.11		32.02	24.62	43.43	36.79	31.38	30.21	22.65	30.33	24.23	33.87	2.08	2.54	8.56
Gemini-2.5 Flash2025.11		29.07	31.1	48.36	38.64	28.36	30.69	19.05	29.04	21.68	34.57	1.38	1.35	8.72
Video-R12025.11		25.56	30.56	47.46	34.58	26.68	29.17	17.13	27.56	18.91	30.33	1.43	1.72	6.45
InternVL3.5-8B2025.11		21.98	23.23	42.26	26.9	22.2	22.45	16.53	23.21	15.64	28.76	0.1	0.08	6.52
VideoChat-R12025.11		21.57	27.47	42.47	30.11	22.56	20.36	15.03	24.7	17.21	25.71	0.64	0.82	6
Qwen2.5-VL-7B-Instruct2025.11		20.27	26.05	38.2	28.43	26.05	22.93	16.01	23.21	16.69	26.56	0.29	0.22	6.07
Mind-Reward only2025.11		18.7	21.84	39.99	27.81	21.35	18.85	21.9	23.3	17.58	24.68	0.28	0.4	6.63
Video-ChatGPT2025.11		9.9	7.04	27	25.72	5.16	16.79	2.7	21.44	1.46	19.95	0	0	5.52
VideoLLaMA3-7B2025.11		7.82	14.8	31.86	30.08	8.09	21.76	4.61	24.28	5.34	19.59	0.63	0.6	5.33
Video-LLaVA2025.11		5.05	2.96	25.33	14.87	6.82	15.55	16.63	15.3	3.29	19.5	0.08	0.08	4.81