List your company

Scientific Agent Task on ScienceAgentBench (test)

18.6Success Rate (SR)SciNav
7.05610.05313.0516.047Mar 11, 2026

Evaluation results

MethodLinks
SciNavBase Model=GPT-4o (2024-11-20), debug steps=32026.0318.669.90.342
SciNavBase Model=GPT-4o (2024-05-13), debug steps=32026.0316.1660.512
Self-DebugBase Model=GPT-4o (2024-11-20), debug steps=102026.0315670.03
Self-DebugBase Model=GPT-4o (2024-05-13), debug steps=102026.0314.771.20.057
OpenHandsBase Model=GPT-4o (2024-05-13)2026.0313.162.81.093
Direct PromptingBase Model=GPT-4o (2024-05-13)2026.037.542.20.011