AI RESEARCH
SWE-Pruner: Self-Adaptive Context Pruning for Coding Agents
arXiv CS.CL
•
ArXi:2601.16746v3 Announce Type: replace-cross LLM agents have nstrated remarkable capabilities in software development, but their performance is hampered by long interaction contexts, which incur high API costs and latency. While various context compression approaches such as LongLLMLingua have emerged to tackle this challenge, they typically rely on fixed metrics such as PPL, ignoring the task-specific nature of code understanding. As a result, they frequently disrupt syntactic and logical structure and fail to retain critical implementation details.