Efficient Code Analysis via Graph Representation Learning-Guided Large Language Models

Hang Gao, Tao Peng, Baoquan Cui, Hong Huang, Fengge Wu, Junsuo Zhao, Jian Zhang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:33429-33459, 2026.

Abstract

Large Language Models (LLMs) have significantly advanced code analysis tasks, yet they struggle to detect malicious behaviors fragmented across files, whose intricate dependencies easily get lost in the vast amount of benign code. We therefore propose a graph-centric attention acquisition pipeline that enhances LLMs’ ability to localize malicious behavior. The approach parses a project into a code graph, uses an LLM to encode nodes with semantic and structural signals, and trains a Graph Neural Network (GNN) under sparse supervision. The GNN performs an initial detection, and by interpreting these predictions, identifies key code sections that are most likely to contain malicious behavior. These influential regions are then used to guide the LLM’s attention for in-depth analysis. This strategy significantly reduces interference from irrelevant context while maintaining low annotation costs. Extensive experiments show that the method consistently outperforms existing approaches on multiple public and custom datasets, highlighting its potential for practical deployment in software security scenarios. Codes can be found in https://github.com/Epiphaniespt/GMLLM.git.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-gao26u, title = {Efficient Code Analysis via Graph Representation Learning-Guided Large Language Models}, author = {Gao, Hang and Peng, Tao and Cui, Baoquan and Huang, Hong and Wu, Fengge and Zhao, Junsuo and Zhang, Jian}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {33429--33459}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/gao26u/gao26u.pdf}, url = {https://proceedings.mlr.press/v306/gao26u.html}, abstract = {Large Language Models (LLMs) have significantly advanced code analysis tasks, yet they struggle to detect malicious behaviors fragmented across files, whose intricate dependencies easily get lost in the vast amount of benign code. We therefore propose a graph-centric attention acquisition pipeline that enhances LLMs’ ability to localize malicious behavior. The approach parses a project into a code graph, uses an LLM to encode nodes with semantic and structural signals, and trains a Graph Neural Network (GNN) under sparse supervision. The GNN performs an initial detection, and by interpreting these predictions, identifies key code sections that are most likely to contain malicious behavior. These influential regions are then used to guide the LLM’s attention for in-depth analysis. This strategy significantly reduces interference from irrelevant context while maintaining low annotation costs. Extensive experiments show that the method consistently outperforms existing approaches on multiple public and custom datasets, highlighting its potential for practical deployment in software security scenarios. Codes can be found in https://github.com/Epiphaniespt/GMLLM.git.} }
Endnote
%0 Conference Paper %T Efficient Code Analysis via Graph Representation Learning-Guided Large Language Models %A Hang Gao %A Tao Peng %A Baoquan Cui %A Hong Huang %A Fengge Wu %A Junsuo Zhao %A Jian Zhang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-gao26u %I PMLR %P 33429--33459 %U https://proceedings.mlr.press/v306/gao26u.html %V 306 %X Large Language Models (LLMs) have significantly advanced code analysis tasks, yet they struggle to detect malicious behaviors fragmented across files, whose intricate dependencies easily get lost in the vast amount of benign code. We therefore propose a graph-centric attention acquisition pipeline that enhances LLMs’ ability to localize malicious behavior. The approach parses a project into a code graph, uses an LLM to encode nodes with semantic and structural signals, and trains a Graph Neural Network (GNN) under sparse supervision. The GNN performs an initial detection, and by interpreting these predictions, identifies key code sections that are most likely to contain malicious behavior. These influential regions are then used to guide the LLM’s attention for in-depth analysis. This strategy significantly reduces interference from irrelevant context while maintaining low annotation costs. Extensive experiments show that the method consistently outperforms existing approaches on multiple public and custom datasets, highlighting its potential for practical deployment in software security scenarios. Codes can be found in https://github.com/Epiphaniespt/GMLLM.git.
APA
Gao, H., Peng, T., Cui, B., Huang, H., Wu, F., Zhao, J. & Zhang, J.. (2026). Efficient Code Analysis via Graph Representation Learning-Guided Large Language Models. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:33429-33459 Available from https://proceedings.mlr.press/v306/gao26u.html.

Related Material