Reasoning and Knowledge Suite (MMLU, ARC, BoolQ, CSQA, HSwag, PIQA, SocIQ, Wino) benchmark leaderboard · SOTA2 Research

Reasoning and Knowledge Suite (MMLU, ARC, BoolQ, CSQA, HSwag, PIQA, SocIQ, Wino)

75.78MMLUQwen3-4B

Evaluation Results

Method	Links
Qwen3-4B2025.12		75.78	89.83	97.53	86.09	81.9	79.46	84.98	75.59	65.43	81.84
Qwen2.5-3B2025.12		66.86	86.44	92.59	83.88	76.09	73.85	81.45	69.4	63.69	77.14
Qwen3-1.7B2025.12		65.35	80.34	91.89	79.82	74.61	60.76	77.2	68.58	59.27	73.09
SmolLM3-3B2025.12		63.04	77.29	88.54	76.12	70.52	69.2	79.05	65.25	64.4	72.6
Qwen2.5-1.5B2025.12		61.56	79.32	90.48	76.39	75.1	64.18	76.17	64.94	59.67	71.98
Llama-3.2-3B2025.12		57.87	72.2	83.95	76.73	70.35	71.06	79.05	64.33	64.09	71.07
Qwen2-1.5B2025.12		56.36	70.17	83.6	71.9	70.52	60.77	75.73	63.46	59.83	68.04
Gemma2-2B2025.12		55.2	66.44	82.54	72.42	69.45	66.2	78.89	65.92	65.35	69.16
Qwen3-0.6B2025.12		55.09	68.14	84.48	69.05	61.18	48.51	69.97	61.51	55.64	63.73
PCMind-2.1-Kaiyuan-2B2025.12		53.9	66.1	82.89	78.53	67.4	58.13	74.37	62.59	65.75	67.74
SmolLM2-1.7B2025.12		51.99	59.66	82.72	69.85	67.16	65.3	78.51	60.18	59.12	66.05
YuLan-Mini-2.4B2025.12		51.76	64.75	82.54	78.59	66.18	61.2	77.31	63.25	61.88	67.5
OLMo-2-0425-1B2025.12		44.25	47.46	76.72	70.55	65.6	61.61	76.44	55.53	60.38	62.06
Llama-3.2-1B2025.12		37.74	36.95	70.55	67.43	62.82	60.2	74.92	50.61	58.17	57.71