List your company

Scientific Reasoning on GPQA Diamond (p@1, p@16)

91.94P@16LED
51.858462.264272.6783.0758Jan 13, 2026Jan 16, 2026Jan 19, 2026Jan 23, 2026Jan 26, 2026Jan 29, 2026Feb 2, 2026

Evaluation results

MethodLinks
LEDModel=Qwen3-30B-A3B-Thinking2026.0291.9471.09
DoLaModel=Qwen3-30B-A3B-Thinking2026.0287.8871.53
ST-GModel=Qwen3-30B-A3B-Thinking2026.0287.8872.06
DoLaModel=Qwen3-4B-Thinking2026.0287.3764.36
DoLaModel=MiMo-7B-RL2026.0286.8751.86
CoTModel=MiMo-7B-RL2026.0286.3651.36
CoTModel=Qwen3-30B-A3B-Thinking2026.0286.3670.77
ST-GModel=MiMo-7B-RL2026.0285.5652.18
ST-GModel=Qwen3-4B-Thinking2026.0285.3564.61
LEDModel=Qwen3-4B-Thinking2026.0285.3564.9
LEDModel=MiMo-7B-RL2026.0285.3552.18
CoTModel=Qwen3-4B-Thinking2026.0283.3363.71
STModel=MiMo-7B-RL2026.0280.351.99
STModel=Qwen3-30B-A3B-Thinking2026.0278.2871.88
STModel=Qwen3-4B-Thinking2026.0277.7864.9
Ministral 3Model Size=14B2026.0171.2—
Qwen3-VLModel Size=8B2026.0167.1—
Ministral 3Model Size=8B2026.0166.8—
Qwen 3Model Size=14B2026.0166.3—
Qwen3-VLModel Size=4B2026.0160.1—
Ministral 3Model Size=3B2026.0153.4—