List your company

Reasoning on GSM8K Rev

86.22AccuracyExGRPO
14.61633.205551.79570.3845Feb 26, 2026

Evaluation results

MethodLinks
ExGRPOStudent Model=Qwen2.5-7B-Instruct2026.0286.22
RevThinkStudent Model=Qwen2.5-7B-Instruct2026.0282.4
AnsAugStudent Model=Qwen2.5-7B-Instruct2026.0279.62
SKDStudent Model=Qwen2.5-7B-Instruct2026.0278.45
Zero-shotStudent Model=Qwen2.5-7B-Instruct2026.0275.1
ExGRPOStudent Model=Gemma-7B-it2026.0234.88
RevThinkStudent Model=Gemma-7B-it2026.0230.12
AnsAugStudent Model=Gemma-7B-it2026.0226.9
SKDStudent Model=Gemma-7B-it2026.0226.05
Zero-shotStudent Model=Gemma-7B-it2026.0217.37