Exploring the Power of Diffusion Large Language Models for Software Engineering: An Empirical Investigation
By: Jingyao Zhang , Tianlin Li , Xiaoyu Zhang and more
Potential Business Impact:
Makes computers write and fix code faster.
Autoregressive Large Language Models (AR-LLMs) are widely used in software engineering (SE) but face limitations in processing code structure information and suffer from high inference latency. Diffusion LLMs (DLLMs) offer a promising alternative with global bidirectional encoding and decoupled generation steps. This work presents the first comprehensive evaluation of DLLMs across the software development lifecycle, including code generation, defect detection, and program repair. On a large-scale benchmark of 52,937 tasks, 7Bparameter DLLMs outperform AR-LLMs with a 30% average accuracy improvement achieving a 113% gain on cross-file repair, while maintaining superior efficiency and reduced latency. Our results establish DLLMs as a superior paradigm for SE tasks.
Similar Papers
Beyond Autoregression: An Empirical Study of Diffusion Large Language Models for Code Generation
Software Engineering
Makes computers write code much faster and better.
Discrete Diffusion in Large Language and Multimodal Models: A Survey
Machine Learning (CS)
Makes AI talk and create much faster.
dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Caching
Machine Learning (CS)
Makes AI text generators work much faster.