Bridging Code Graphs and Large Language Models for Better Code Understanding

TL;DR

This paper introduces CGBridge, a plug-and-play method that enhances large language models with code graph information via an external trainable module, significantly improving code understanding tasks without modifying the LLM architecture.

Contribution

It proposes a novel external bridge module trained on large-scale code graphs to incorporate structural semantics into LLMs for better code understanding.

Findings

Achieves 16.19% and 9.12% improvements in code summarization accuracy.

Yields 9.84% and 38.87% gains in code translation accuracy.

Over 4x faster inference compared to LoRA-tuned models.

Abstract

Large Language Models (LLMs) have demonstrated remarkable performance in code intelligence tasks such as code generation, summarization, and translation. However, their reliance on linearized token sequences limits their ability to understand the structural semantics of programs. While prior studies have explored graphaugmented prompting and structure-aware pretraining, they either suffer from prompt length constraints or require task-specific architectural changes that are incompatible with large-scale instructionfollowing LLMs. To address these limitations, this paper proposes CGBridge, a novel plug-and-play method that enhances LLMs with Code Graph information through an external, trainable Bridge module. CGBridge first pre-trains a code graph encoder via selfsupervised learning on a large-scale dataset of 270K code graphs to learn structural code semantics. It then trains an external module to bridge the modality gap among code, graph, and text by aligning their semantics through cross-modal attention mechanisms. Finally, the bridge module generates structure-informed prompts, which are injected into a frozen LLM, and is fine-tuned for downstream code intelligence tasks. Experiments show that CGBridge achieves notable improvements over both the original model and the graphaugmented prompting method. Specifically, it yields a 16.19% and 9.12% relative gain in LLM-as-a-Judge on code summarization, and a 9.84% and 38.87% relative gain in Execution Accuracy on code translation. Moreover, CGBridge achieves over 4x faster inference than LoRA-tuned models, demonstrating both effectiveness and efficiency in structure-aware code understanding.