List your company

Software Engineering Issue Resolution on Terminal Bench

32.5Resolve RateHybrid
14.319.02523.7528.475Dec 26, 2025

Evaluation results

MethodLinks
HybridFeedback=Hybrid (Execution-free + Execution-based)2025.1232.5
Execution-free onlyFeedback=Execution-free (SWE-RM)2025.1231.3
Execution-based onlyFeedback=Execution-based2025.1230
Poor Calibrated RMFeedback=Ablated RM (Poorly Calibrated)2025.1215