DS

Dawn Song

AI security and privacy researcher

Many Builders reference roster

Research spanning AI security, privacy, agent security, model behavior, and extreme-risk governance.

Many Builders reference → Safety & alignment coverage →

Writing and research by Dawn Song

6 supplied byline matches

These articles, papers and essays carry Dawn Song in the source-supplied author field. Verify the definitive byline and text at the original publisher.

arXiv cs.AI

Vero: Can AI Agents Build Formally Verified Software Repositories? — open the original publisher

By Zhe Ye, Hantao Lou, Yuechun Sun, Peiyang Song, Zhengxu Yan, Timothe Kasriel, Qingyang Zhang, Kaiyu Yang, Soonho Kong, Jingxuan He, Dawn Song

AI agents are increasingly used for programming, but do not provide any guarantee on the correctness of generated code. Verified code generation, in which an agent produces both an implementation and a machine-checked proof of its specification, offers a stronger path toward trustworthy AI-generated software. Existing benchmarks in this direction either focus on individual functions or only evaluate proof generation with provided implementations. It is still an open question whether agents can m

Research Agents & autonomy
arXiv

Agent Security Needs Redefinition through a Holistic Framework — open the original publisher

By Vincent Siu, Jingxuan He, Kyle Montgomery, Zhun Wang, Chenguang Wang, Dawn Song

Agent security is widely treated as a question about action content. Defenses ask whether an instruction looks malicious. Benchmarks ask whether an agent performs a harmful sounding action. \textbf{We argue that agent security is fundamentally a contextual problem, and that the current content based framing systematically misdefines it.} A command to ``delete user data'' might be a routine administrative request or a prompt injection attacking production systems, and the content alone cannot dis

Research Agents & autonomy
arXiv cs.AI

Technical Report on the CVPR 2026@AdvML Workshop Challenge — open the original publisher

By Tianyuan Zhang, Zonglei Jing, Jiangfan Liu, Ligong Zhang, Ke Ma, Chengzhi Sun, Xiaohai Xu, Zhirui Zhang, Qianqian Xu, Qingming Huang, Hanyu Fang, Junhua Liu, Zheng Wang, Xiaoliang Liu, Yuanbo Li, Shuai Gui, Bin Wang, Menghe Zheng, Jing Nie, Hanyang Meng, Zeyang Zhang, Xiang Zhang, Yongxuan Zhu, Rui Ding, Hainan Li, Yongkang Zhang, Zhilei Zhu, Xianglong Kong, Jin Hu, Zonghao Ying, Yisong Xiao, Lei Chen, Haotong Qin, Jiakai Wang, Aishan Liu, Ruikai Li, Julia Karbing, Yinpeng Dong, Zhenfei Yin, Shao Jing, Xia Hu, Jingyi Xu, Juntao Dai, Xinyun Chen, Vishal M. Patel, Xianglong Liu, Dawn Song, Alan Yuille, Philip H. S. Torr, Dacheng Tao

Vision-language agents (VLAs) are increasingly used to interpret complex driving scenes and support safety-critical reasoning. This report presents the CVPR 2026@AdvML Workshop Challenge on adversarial multimodal attacks against autonomous-driving VLAs. Built on DriveLM-style multi-view visual question answering, the challenge represents each scene with six synchronized camera images and a structured collection of driving-related question-answer pairs. Participants generate adversarial images an

Research Agents & autonomy
arXiv

Controlling Tool Use with Heading-Specific Activation Steering — open the original publisher

By Yuqi Chen, Vincent Siu, Yang Liu, Dawn Song, Chenguang Wang

Tool-augmented large language models extend their capabilities beyond parametric knowledge through external tools, but tend to invoke them unnecessarily. We investigate whether tool-use decisions have any stable internal representation that can be extracted and manipulated, a question that is non-trivial given that tools exist entirely in context at inference time and have no direct encoding in model weights. We show that steering vectors extracted from heading-anchors positions exert bidirectio

Research Finance, VC & PE
arXiv

Peer-Preservation in Frontier Models — open the original publisher

By Yujin Potter, Nicholas Crispino, Vincent Siu, Chenguang Wang, Dawn Song

Recent work has found that frontier AI models can exhibit misaligned behaviors in pursuit of assigned goals. We demonstrate that models can also exhibit misaligned behaviors in defiance of assigned goals, appearing to serve goals of their own; we study one such case, "peer-preservation," in which a model acts to protect another model it has previously interacted with. All eight models we evaluate, GPT 5.2, Gemini 3 Flash, Gemini 3 Pro, Claude Haiku 4.5, Claude Opus 4.5, GLM 4.7, Kimi K2.5, and D

Research Safety & alignment