{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T20:30:50Z","timestamp":1792787450765,"version":"4.5.0"},"reference-count":52,"publisher":"Association for Computing Machinery (ACM)","issue":"7","funder":[{"name":"National Key Research and Development Program of China","award":["2022YFB4501902"],"award-info":[{"award-number":["2022YFB4501902"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Softw. Eng. Methodol."],"published-print":{"date-parts":[[2026,7,31]]},"abstract":"<jats:p>Large Language Models (LLMs) have strong capabilities in code comprehension, but fine-tuning costs and semantic alignment issues limit their project-specific optimization; conversely, fine-tuned models such as CodeBERT are easy to fine-tune, but it is often difficult to learn vulnerability semantics from complex code languages. To address these challenges, this article introduces the Multi-Model Collaborative Vulnerability Detection approach (M2CVD) that leverages the strong capability of analyzing vulnerability semantics from LLMs to improve the detection accuracy of fine-tuned models. M2CVD employs a novel collaborative process: first enhancing the quality of vulnerability description produced by LLMs through the understanding of project code by fine-tuned models, and then using these improved vulnerability descriptions to boost the detection accuracy of fine-tuned models. M2CVD include three main phases: (1) Initial Vulnerability Detection: The initial vulnerability detection is conducted by fine-tuning a detection model (e.g., CodeBERT) and interacting with an LLM (e.g., ChatGPT), respectively. The vulnerability description will be generated by the LLM when the code is detected vulnerable by the LLM; (2) Vulnerability Description Refinement: By informing the LLM of the vulnerability assessment results of the detection model, we refine the vulnerability description by interacting with the LLM. Such refinement can enhance LLM\u2019s vulnerability understanding in specific projects, effectively bridging the previously mentioned alignment gap; (3) Integrated Vulnerability Detection: M2CVD integrates code fragment and the refined vulnerability descriptions inferred to form synthetic data. Then, the synthetic data is used to fine-tune a validation model, optimize the defect feature learning efficiency of the model, and improve the detection accuracy. We demonstrated M2CVD\u2019s effectiveness on two real-world datasets, where M2CVD significantly outperformed the baseline. In addition, we demonstrate that the M2CVD collaborative method can extend to other different LLMs and fine-tuned models to improve their accuracy in vulnerability detection tasks.<\/jats:p>","DOI":"10.1145\/3771923","type":"journal-article","created":{"date-parts":[[2025,10,16]],"date-time":"2025-10-16T14:45:01Z","timestamp":1760625901000},"page":"1-25","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":6,"title":["M2CVD: Enhancing Vulnerability Understanding through Multi-Model Collaboration for Code Vulnerability Detection"],"prefix":"10.1145","volume":"35","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7534-0059","authenticated-orcid":false,"given":"Ziliang","family":"Wang","sequence":"first","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5828-0186","authenticated-orcid":false,"given":"Ge","family":"Li","sequence":"additional","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5579-8852","authenticated-orcid":false,"given":"Jia","family":"Li \u2642","sequence":"additional","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9411-971X","authenticated-orcid":false,"given":"Jia","family":"Li","sequence":"additional","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9538-9121","authenticated-orcid":false,"given":"Meng","family":"Yan","sequence":"additional","affiliation":[{"name":"School of Big Data &amp; Software Engineering, Chongqing University, Chongqing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8991-747X","authenticated-orcid":false,"given":"Yingfei","family":"Xiong","sequence":"additional","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1087-226X","authenticated-orcid":false,"given":"Zhi","family":"Jin","sequence":"additional","affiliation":[{"name":"Key Lab of High Confidence Software Technology, MoE, School of Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,7,10]]},"reference":[{"key":"e_1_3_2_2_2","unstructured":"An Yang Baosong Yang Binyuan Hui Bo Zheng Bowen Yu Chang Zhou Chengpeng Li Chengyuan Li Dayiheng Liu Fei Huang et al. 2024. Qwen2 Technical Report. arXiv:2407.10671. Retrieved from https:\/\/arxiv.org\/abs\/2407.10671"},{"key":"e_1_3_2_3_2","first-page":"3011","volume-title":"Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (Main Volume)","author":"Bai Jiangang","year":"2021","unstructured":"Jiangang Bai, Yujing Wang, Yiren Chen, Yaming Yang, Jing Bai, Jing Yu, and Yunhai Tong. 2021. Syntax-BERT: Improving pre-trained transformers with syntax trees. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (Main Volume), 3011\u20133020."},{"key":"e_1_3_2_4_2","doi-asserted-by":"publisher","DOI":"10.1145\/3510003.3510219"},{"key":"e_1_3_2_5_2","doi-asserted-by":"crossref","first-page":"3280","DOI":"10.1109\/TSE.2021.3087402","article-title":"Deep learning based vulnerability detection: Are we there yet","volume":"48","author":"Chakraborty Saikat","year":"2021","unstructured":"Saikat Chakraborty, Rahul Krishna, Yangruibo Ding, and Baishakhi Ray. 2021. Deep learning based vulnerability detection: Are we there yet. IEEE Transactions on Software Engineering 48 (2021), 3280\u20133296.","journal-title":"IEEE Transactions on Software Engineering"},{"key":"e_1_3_2_6_2","doi-asserted-by":"publisher","DOI":"10.5555\/1622407.1622416"},{"key":"e_1_3_2_7_2","unstructured":"Checkmarx. 2022. CheckmarX. Retrieved from https:\/\/www.checkmarx.com\/."},{"key":"e_1_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.1145\/3436877"},{"key":"e_1_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1145\/3533767.3534371"},{"key":"e_1_3_2_10_2","unstructured":"Hoa Khanh Dam Truyen Tran Trang Pham Shien Wee Ng John Grundy and Aditya Ghose. 2017. Automatic feature learning for vulnerability prediction. arXiv:1708.02368. Retrieved from https:\/\/arxiv.org\/abs\/1708.02368"},{"key":"e_1_3_2_11_2","unstructured":"DeepSeek-AI. 2024. DeepSeek-V2: A strong economical and efficient mixture-of-experts language model. arXiv:2405.04434. Retrieved from https:\/\/arxiv.org\/abs\/2405.04434"},{"key":"e_1_3_2_12_2","first-page":"1","volume-title":"Proceedings of the 46th IEEE\/ACM International Conference on Software Engineering ICSE 2024","author":"Ding Yangruibo","year":"2024","unstructured":"Yangruibo Ding, Benjamin Steenhoek, Kexin Pei, Gail Kaiser, Wei Le, and Baishakhi Ray. 2024. Traced: Execution-aware pre-training for source code. In Proceedings of the 46th IEEE\/ACM International Conference on Software Engineering ICSE 2024, 1\u201312."},{"key":"e_1_3_2_13_2","first-page":"4665","volume-title":"Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI \u201919)","author":"Duan Xu","year":"2019","unstructured":"Xu Duan, Jingzheng Wu, Shouling Ji, Zhiqing Rui, Tianyue Luo, Mutian Yang, and Yanjun Wu. 2019. VulSniper: Focus your attention to shoot fine-grained vulnerabilities. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI \u201919), 4665\u20134671."},{"key":"e_1_3_2_14_2","doi-asserted-by":"crossref","first-page":"1536","DOI":"10.18653\/v1\/2020.findings-emnlp.139","volume-title":"Proceedings of the Findings of the Association for Computational Linguistics (EMNLP \u201920)","author":"Feng Zhangyin","year":"2020","unstructured":"Zhangyin Feng, Daya Guo, Duyu Tang, Nan Duan, Xiaocheng Feng, Ming Gong, Linjun Shou, Bing Qin, Ting Liu, Daxin Jiang, et al. 2020. CodeBERT: A pre-trained model for programming and natural languages. In Proceedings of the Findings of the Association for Computational Linguistics (EMNLP \u201920), 1536\u20131547."},{"key":"e_1_3_2_15_2","unstructured":"Flawfinder. 2022. Flawfinder. Retrieved from https:\/\/dwheeler.com\/flawfinder\/"},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3524842.3528452"},{"key":"e_1_3_2_17_2","unstructured":"Michael Fu Chakkrit Tantithamthavorn Van Nguyen and Trung Le. 2023. ChatGPT for vulnerability detection classification and repair: how far are we? arXiv:2310.09810. Retrieved from https:\/\/arxiv.org\/abs\/2310.09810"},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.499"},{"key":"e_1_3_2_19_2","first-page":"1","volume-title":"Proceedings of the International Conference on Learning Representations\u2013","author":"Guo Daya","year":"2020","unstructured":"Daya Guo, Shuo Ren, Shuai Lu, Zhangyin Feng, Duyu Tang, L. I. U. Shujie, Long Zhou, Nan Duan, Alexey Svyatkovskiy, Shengyu Fu, et al. 2020. GraphCodeBERT: Pre-training code representations with data flow. In Proceedings of the International Conference on Learning Representations, 1\u201318."},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3524842.3527949"},{"key":"e_1_3_2_21_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcss.2014.02.005"},{"issue":"128","key":"e_1_3_2_22_2","first-page":"16","article-title":"Guide for security-focused configuration management of information systems","volume":"800","author":"Johnson Arnold","year":"2011","unstructured":"Arnold Johnson, Kelley Dempsey, Ron Ross, Sarbari Gupta, and Dennis Bailey. 2011. Guide for security-focused configuration management of information systems. NIST Special Publication 800, 128 (2011), 16\u201316.","journal-title":"NIST Special Publication"},{"key":"e_1_3_2_23_2","first-page":"5110","volume-title":"International Conference on Machine Learning","author":"Kanade Aditya","year":"2020","unstructured":"Aditya Kanade, Petros Maniatis, Gogul Balakrishnan, and Kensen Shi. 2020. Learning and evaluating contextual embedding of source code. In International Conference on Machine Learning. PMLR, 5110\u20135121."},{"key":"e_1_3_2_24_2","first-page":"14967","volume-title":"Advances in Neural Information Processing Systems","volume":"34","author":"Lachaux Marie-Anne","year":"2021","unstructured":"Marie-Anne Lachaux, Baptiste Roziere, Marc Szafraniec, and Guillaume Lample. 2021. DOBF: A deobfuscation pre-training objective for programming languages. In Advances in Neural Information Processing Systems, Vol. 34, 14967\u201314979."},{"key":"e_1_3_2_25_2","doi-asserted-by":"publisher","DOI":"10.1145\/3468264.3468597"},{"key":"e_1_3_2_26_2","volume-title":"Proceedings of the International Conference on Learning Representations (ICLR \u201916)","author":"Li Yujia","year":"2016","unstructured":"Yujia Li, Richard Zemel, Marc Brockschmidt, and Daniel Tarlow. 2016. Gated graph sequence neural networks. In Proceedings of the International Conference on Learning Representations (ICLR \u201916)."},{"key":"e_1_3_2_27_2","doi-asserted-by":"publisher","DOI":"10.1109\/TDSC.2021.3051525"},{"key":"e_1_3_2_28_2","unstructured":"Zhen Li Deqing Zou Shouhuai Xu Xinyu Ou Hai Jin Sujuan Wang Zhijun Deng and Yuyi Zhong. 2018. Vuldeepecker: A deep learning-based system for vulnerability detection. arXiv:1801.01681. Retrieved from https:\/\/arxiv.org\/abs\/1801.01681"},{"key":"e_1_3_2_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3133956.3138840"},{"key":"e_1_3_2_30_2","first-page":"324","volume-title":"Proceedings of the 2021 IEEE\/ACM 43rd International Conference on Software Engineering (ICSE)","author":"Lin Jinfeng","year":"2021","unstructured":"Jinfeng Lin, Yalin Liu, Qingkai Zeng, Meng Jiang, and Jane Cleland-Huang. 2021. Traceability transformed: Generating more accurate links with pre-trained BERT models. In Proceedings of the 2021 IEEE\/ACM 43rd International Conference on Software Engineering (ICSE). IEEE, 324\u2013335."},{"key":"e_1_3_2_31_2","first-page":"1","volume-title":"Proceedings of the IEEE\/ACM 46th International Conference on Software Engineering","author":"Liu Zhongxin","year":"2024","unstructured":"Zhongxin Liu, Zhijie Tang, Junwei Zhang, Xin Xia, and Xiaohu Yang. 2024. Pre-training by predicting program dependencies for vulnerability analysis tasks. In Proceedings of the IEEE\/ACM 46th International Conference on Software Engineering, 1\u201313."},{"key":"e_1_3_2_32_2","unstructured":"Llama. 2022. Meta AI Publications. Retrieved from https:\/\/ai.meta.com\/llama\/."},{"key":"e_1_3_2_33_2","first-page":"1","volume-title":"Advances in Neural Information Processing Systems","volume":"32","author":"Mao Chengzhi","year":"2019","unstructured":"Chengzhi Mao, Ziyuan Zhong, Junfeng Yang, Carl Vondrick, and Baishakhi Ray. 2019. Metric learning for adversarial robustness. In Advances in Neural Information Processing Systems, Vol. 32, 1\u201312."},{"key":"e_1_3_2_34_2","unstructured":"Qingkai Min Qipeng Guo Xiangkun Hu Songfang Huang Zheng Zhang and Yue Zhang. 2024. Synergetic event understanding: A collaborative approach to cross-document event coreference resolution with large language models. arXiv:2406.02148. Retrieved from https:\/\/arxiv.org\/abs\/2406.02148"},{"key":"e_1_3_2_35_2","doi-asserted-by":"publisher","DOI":"10.1145\/3510454.3516865"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1145\/3510003.3510096"},{"key":"e_1_3_2_37_2","unstructured":"ChatGPT. 2022. openAI. Retrieved from https:\/\/www.chat.openai.com\/."},{"key":"e_1_3_2_38_2","doi-asserted-by":"crossref","first-page":"1687","DOI":"10.1145\/3412841.3442132","volume-title":"Proceedings of the 36th Annual ACM Symposium on Applied Computing","author":"Rabheru Rishi","year":"2021","unstructured":"Rishi Rabheru, Hazim Hanif, and Sergio Maffeis. 2021. DeepTective: Detection of PHP vulnerabilities using hybrid graph neural networks. In Proceedings of the 36th Annual ACM Symposium on Applied Computing, 1687\u20131690."},{"key":"e_1_3_2_39_2","unstructured":"Baptiste Roziere Jonas Gehring Fabian Gloeckle Sten Sootla Itai Gat Xiaoqing Ellen Tan Yossi Adi Jingyu Liu Tal Remez J\u00e9r\u00e9my Rapin et al. 2023. Code llama: Open foundation models for code. arXiv:2308.12950. Retrieved from https:\/\/arxiv.org\/abs\/2308.12950."},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICMLA.2018.00120"},{"key":"e_1_3_2_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/78.650093"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICSE48619.2023.00188"},{"key":"e_1_3_2_43_2","unstructured":"Benjamin Steenhoek Md Mahbubur Rahman Shaila Sharmin and Wei Le. 2023. Do language models learn semantics of code? A case study in vulnerability detection. arXiv:2311.04109. Retrieved from https:\/\/arxiv.org\/abs\/231104109"},{"key":"e_1_3_2_44_2","doi-asserted-by":"publisher","DOI":"10.1145\/3564625.3567985"},{"key":"e_1_3_2_45_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.685"},{"key":"e_1_3_2_46_2","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1800"},{"key":"e_1_3_2_47_2","unstructured":"Xin-Cheng Wen Yupan Chen Cuiyun Gao Hongyu Zhang Jie M. Zhang and Qing Liao. 2023. Vulnerability detection with graph simplification and enhanced graph representation learning. arXiv:2302.04675. Retrieved from https:\/\/arxiv.org\/abs\/2302.04675"},{"key":"e_1_3_2_48_2","doi-asserted-by":"publisher","DOI":"10.1145\/3510003.3510229"},{"key":"e_1_3_2_49_2","volume-title":"Pattern-Based Vulnerability Discovery","author":"Fabian Yamaguchi","year":"2015","unstructured":"Yamaguchi Fabian. 2015. Pattern-Based Vulnerability Discovery. Ph.D. Dissertation. University of G\u00f6ttingen."},{"issue":"2","key":"e_1_3_2_50_2","first-page":"101","article-title":"Pattern-based methods for vulnerability discovery","volume":"59","author":"Yamaguchi Fabian","year":"2017","unstructured":"Fabian Yamaguchi. 2017. Pattern-based methods for vulnerability discovery. Information Technology 59, 2 (2017), 101\u2013106.","journal-title":"Information Technology"},{"issue":"8","key":"e_1_3_2_51_2","article-title":"Vulnerability detection by learning from syntax-based execution paths of code","volume":"49","author":"Zhang Junwei","year":"2023","unstructured":"Junwei Zhang, Zhongxin Liu, Xing Hu, Xin Xia, and Shanping Li. 2023. Vulnerability detection by learning from syntax-based execution paths of code. IEEE Transactions on Software Engineering 49, 8 (2023), 4196\u20134212.","journal-title":"IEEE Transactions on Software Engineering"},{"key":"e_1_3_2_52_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISSRE52982.2021.00054"},{"key":"e_1_3_2_53_2","first-page":"1","volume-title":"Advances in Neural Information Processing Systems","volume":"32","author":"Zhou Yaqin","year":"2019","unstructured":"Yaqin Zhou, Shangqing Liu, Jingkai Siow, Xiaoning Du, and Yang Liu. 2019. Devign: Effective vulnerability identification by learning comprehensive program semantics via graph neural networks. In Advances in Neural Information Processing Systems, Vol. 32s, 1\u201311."}],"container-title":["ACM Transactions on Software Engineering and Methodology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3771923","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T14:42:45Z","timestamp":1783694565000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3771923"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,10]]},"references-count":52,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2026,7,31]]}},"alternative-id":["10.1145\/3771923"],"URL":"https:\/\/doi.org\/10.1145\/3771923","relation":{},"ISSN":["1049-331X","1557-7392"],"issn-type":[{"value":"1049-331X","type":"print"},{"value":"1557-7392","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,10]]},"assertion":[{"value":"2024-07-19","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-10-09","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-07-10","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}