SWE-Compass: Towards Unified Evaluation of Agentic Coding Abilities for Large Language Models

Open in new window