The SWE-bench [1] evaluation framework has catalyzed the development of multi-agent large language model (LLM) systems for addressing real-world software engineering tasks, with an initial focus on ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback