SWE-QA

The SWE-QA dataset/benchmark contains a collection of software engineering questions designed to evaluate the performance of coding agents in understanding and generating code-related responses.

Papers using SWE-QA (3)

FastCode: Fast and Cost-Efficient Code Understanding and Reasoning2026SWE-QA: Can Language Models Answer Repository-level Code Questions?2025FastContext: Training Efficient Repository Explorer for Coding Agents2026