List your company

Sequential Decision Making on ScienceWorld

70.4Pass@1 Success RateAutoRefine
61.45663.77866.168.422Jan 30, 2026

Evaluation results

MethodLinks
AutoRefineBackbone=GPT-4-turbo2026.0170.416.5
ReAct + ReflexionBackbone=GPT-4-turbo2026.0169.240.2
ReflexionBackbone=GPT-4-turbo2026.0167.442
ReActBackbone=GPT-4-turbo2026.0161.826.2