MutaGReP: Execution-Free Repository-Grounded Plan Search for Code-Use
Zaid Khan, Ali Farhadi, Ranjay Krishna, Luca Weihs, Mohit Bansal,, Tanmay Gupta

TL;DR
MutaGReP introduces a novel plan search method that decomposes coding tasks into natural language steps grounded in code repositories, enabling large language models to perform effectively with minimal context.
Contribution
It presents a mutation-guided neural tree search approach for plan generation that reduces context requirements while maintaining high coding performance.
Findings
Plans use less than 5% of the 128K context window.
MutaGReP enables models to match GPT-4o performance with full repo context.
Progress on the hardest LongCodeArena tasks.
Abstract
When a human requests an LLM to complete a coding task using functionality from a large code repository, how do we provide context from the repo to the LLM? One approach is to add the entire repo to the LLM's context window. However, most tasks involve only fraction of symbols from a repo, longer contexts are detrimental to the LLM's reasoning abilities, and context windows are not unlimited. Alternatively, we could emulate the human ability to navigate a large repo, pick out the right functionality, and form a plan to solve the task. We propose MutaGReP (Mutation-guided Grounded Repository Plan Search), an approach to search for plans that decompose a user request into natural language steps grounded in the codebase. MutaGReP performs neural tree search in plan space, exploring by mutating plans and using a symbol retriever for grounding. On the challenging LongCodeArena benchmark, our…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsSoftware Testing and Debugging Techniques · Natural Language Processing Techniques · Software Engineering Research
