List your company

Grade School Mathematical Reasoning on GSM8K (Pass@1, #Tokens)

95.88Pass@1MixReasoning
83.71286.87190.0393.189Oct 7, 2025

Evaluation results

MethodLinks
MixReasoningType=ADAPTIVE2025.1095.881,389
BeConciseType=SHORT COT2025.1095.681,822
DEERType=SHORT COT2025.1095.621,223
O1-PrunerType=SHORT COT2025.1095.561,323
Qwen3-8BType=BASE LLM2025.1095.452,403
NoWaitType=SHORT COT2025.1095.381,406
CoT-ValveType=SHORT COT2025.1088.46784
ThinklessType=HYBRID2025.1084.18624