Statically Contextualizing Large Language Models with Typed Holes
This addresses code completion accuracy for developers working with application-specific data structures, though it is incremental as it builds on existing IDE and language server infrastructure.
The paper tackles the problem of LLM-based code completion systems hallucinating broken code due to lack of context, by integrating LLMs with language server type and binding structures to provide token-efficient contextualization. Results show that contextualization with type definitions is particularly impactful, validated through MVUBench dataset tests and porting to TypeScript.
Large language models (LLMs) have reshaped the landscape of program synthesis. However, contemporary LLM-based code completion systems often hallucinate broken code because they lack appropriate context, particularly when working with definitions not in the training data nor near the cursor. This paper demonstrates that tight integration with the type and binding structure of a language, as exposed by its language server, can address this contextualization problem in a token-efficient manner. In short, we contend that AIs need IDEs, too! In particular, we integrate LLM code generation into the Hazel live program sketching environment. The Hazel Language Server identifies the type and typing context of the hole being filled, even in the presence of errors, ensuring that a meaningful program sketch is always available. This allows prompting with codebase-wide contextual information not lexically local to the cursor, nor necessarily in the same file, but that is likely to be semantically local to the developer's goal. Completions synthesized by the LLM are then iteratively refined via further dialog with the language server. To evaluate these techniques, we introduce MVUBench, a dataset of model-view-update (MVU) web applications. These applications serve as challenge problems due to their reliance on application-specific data structures. We find that contextualization with type definitions is particularly impactful. After introducing our ideas in the context of Hazel we duplicate our techniques and port MVUBench to TypeScript in order to validate the applicability of these methods to higher-resource languages. Finally, we outline ChatLSP, a conservative extension to the Language Server Protocol (LSP) that language servers can implement to expose capabilities that AI code completion systems of various designs can use to incorporate static context when generating prompts for an LLM.