CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback
This addresses the need for scalable, validated synthetic data for code generation in AI, though it is incremental as it builds on existing synthesis methods with improved validation.
The paper tackles the problem of acquiring high-quality instruction-code pairs for training LLMs in code generation by proposing CodeEvo, a framework that synthesizes data through iterative interactions between two LLM agents and a hybrid feedback mechanism, resulting in models fine-tuned on this data significantly outperforming baselines across benchmarks.
Acquiring high-quality instruction-code pairs is essential for training Large Language Models (LLMs) for code generation. Manually curated data is expensive and inherently limited in scale, motivating the development of code-centric synthesis methods. Yet, current approaches either focus on augmenting existing code or rely on predefined heuristics, both lacking rigorous data validation, which results in synthetic data that is ungrounded, repetitive, or overly simplistic. Inspired by collaborative programming practices, we propose CodeEvo, a framework that synthesizes code data through iterative interactions between two LLM agents: a Coder, which generates candidate code and test cases based on given instructions, and a Reviewer, which guides the synthesis process by producing new instructions and feedback. We further introduce a hybrid feedback mechanism that combines compiler determinism with the generative flexibility of agents, enabling automatic quality control throughout synthesis. Extensive experiments demonstrate that models fine-tuned on CodeEvo data significantly outperform established baselines across code generation benchmarks with various difficulties. In-depth analyses further provide insights from multiple perspectives into effective code-centric data synthesis.