ACL Findings2025
Boosting Vulnerability Detection of LLMs via Curriculum Preference Optimization with Synthetic Reasoning Data
Xin-Cheng Wen, Yijun Yang, Cuiyun Gao, Yang Xiao, Deheng Ye
Abstract
Large language models (LLMs) demonstrate considerable proficiency in numerous codingrelated tasks; however, their capabilities in detecting software vulnerabilities remain limited. This limitation primarily stems from two factors: (1) the absence of reasoning data related to vulnerabilities, which hinders the models' ability to capture underlying vulnerability patterns; and (2) their focus on learning semantic representations rather than the reason behind them, thus failing to recognize semantically similar vulnerability samples. Furthermore, the development of LLMs specialized in vulnerability detection is challenging, particularly in environments characterized by the scarcity of high-quality datasets. In this paper, we propose a novel framework ReVD that excels at mining vulnerability patterns through reasoning data synthesizing and vulnerability-specific preference optimization. Specifically, we construct forward and backward reasoning processes for vulnerability and corresponding fixed code, ensuring the synthesis of high-quality reasoning data. Moreover, we design the triplet supervised fine-tuning followed by curriculum online preference optimization for enabling ReVD to better understand vulnerability patterns. The extensive experiments conducted on PrimeVul and SVEN datasets demonstrate that ReVD sets new state-of-the-art for LLM-based software vulnerability detection, e.g., 12.24%-22.77% improvement in the accuracy. The source code and data are available at https: //github.com/Xin-Cheng-Wen/PO4Vul.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
4 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation cites Direct Preference Optimization: Your Language Model is Secretly a Reward Model
- Vul-R2: A Reasoning LLM for Automated Vulnerability Repair cites Data Quality for Software Vulnerability Datasets
- Repo2Run: Automated Building Executable Environment for Code Repository at Scale cites Vul-R2: A Reasoning LLM for Automated Vulnerability Repair
- SEER: Enhancing Chain-of-Thought Code Generation through Self-Exploring Deep Reasoning cites Vul-R2: A Reasoning LLM for Automated Vulnerability Repair
- Data Quality for Software Vulnerability Datasets cites CodeBERT: A Pre-Trained Model for Programming and Natural Languages
Cited by top-tier papers4
- Repo2Run: Automated Building Executable Environment for Code Repository at ScaleRuida Hu, Chao Peng, Xinchen Wang, Junjielong Xu et al.NeurIPS 2025 · 49 citations
- Vul-R2: A Reasoning LLM for Automated Vulnerability RepairXin-Cheng Wen, Zirui Lin, Yijun Yang, Cuiyun Gao et al.ASE 2025 · 1 citation
- SEER: Enhancing Chain-of-Thought Code Generation through Self-Exploring Deep ReasoningShuzheng Gao, Chaozheng Wang, Cuiyun Gao, Michael R. LyuICSE 2026 · 1 citation
- VulInstruct: Teaching LLMs Root-Cause Reasoning for Vulnerability Detection via Security SpecificationsHao Zhu, Jia Li, Cuiyun Gao, Jiaru Qian et al.FSE 2026
Builds on14
- Direct Preference Optimization: Your Language Model is Secretly a Reward ModelRafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D. Manning et al.NeurIPS 2023 · 10,924 citations
- CodeBERT: A Pre-Trained Model for Programming and Natural LanguagesZhangyin Feng, Daya Guo, Duyu Tang, Nan Duan et al.EMNLP Findings 2020 · 2,828 citations
- Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine TranslationHaoran Xu, Amr Sharaf, Yunmo Chen, Weiting Tan et al.ICML 2024 · 447 citations
- Enhancing Static Analysis for Practical Bug Detection: An LLM-Integrated ApproachHaonan Li, Yu Hao, Yizhuo Zhai, Zhiyun QianOOPSLA 2024 · 142 citations
- Data Quality for Software Vulnerability DatasetsRoland Croft, Muhammad Ali Babar, M. Mehdi KholoosiICSE 2023 · 138 citations
Related papers
- Three Heads Are Better Than One: A Multi-perspective Reasoning Framework for Enhanced Vulnerability DetectionXin Peng, Bo Lin, Jing Wang, Xiaoling Li et al.FSE 2026 · 1 citation
- RealVul: Can We Detect Vulnerabilities in Web Applications with LLM?Di Cao, Yong Liao, Xiuwei ShangEMNLP 2024 · 11 citations
- Vulnerability Detection with Code Language Models: How Far are We?Yangruibo Ding, Yanjun Fu, Omniyyah Ibrahim, Chawin Sitawarin et al.ICSE 2025 · 44 citations
- Enhancing Vulnerability Detection via Inter-procedural Semantic CompletionBozhi Wu, Chengjie Liu, Zhiming Li, Yushi Cao et al.ISSTA 2025 · 2 citations
- Thought Is All You Need: Smart Contract Vulnerability Detection with Thought-Augmented Large Language ModelChaoyuan Peng, Muhui Jiang, Yajin Zhou, Lei WuFSE 2026
