Hao Wang

Ph.D. Student@UCB

avatar.jpg

I am a second-year Ph.D. student in computer science at University of California, Berkeley, jointly advised by Koushik Sen and Dawn Song. I also collaborate closely with David Wagner. My interest lies generally in the intersection of AI and Security.

During my undergraduate, I worked on software testing with Darko Marinov, Tao Xie, and Tianyin Xu. I also worked on quantum computing with Tongyang Li with a focus on optimization algorithms. You can find my CV here: Tony’s Curriculum Vitae

news

Oct 1, 2025 I am honored to be selected as an Amazon AI Fellow! :tada:
Apr 15, 2025 Our paper Type-Constrained Code Generation with Language Models is accepted at PLDI 2025!
Sep 2, 2023 I am selected as a young researcher at HLF 2023! Looking forward to meeting the laureates and peer researchers :laughing: :satisfied:

selected publications

  1. AutoSWE
    Towards Autonomous Software Engineering
    Wang, Hao, Meng, Ruijie, Ye, Zhe, Gu, Alex, Jain, Naman, Liu, Xiaoyuan, Chaudhuri, Swarat, Zimmermann, Thomas, Gulwani, Sumit, Solar-Lezama, Armando, Stoica, Ion, and Song, Dawn
    2026
    PDF
  2. ConfFuzz
    ConfFuzz: Parameter-Aware Greybox Fuzzing for Configurable Cloud Systems
    Wang, Shuai,  Wang, Hao, Marinov, Darko, Xu, Tianyin, and Zhang, Yongle
    In ASE 2026
  3. BenchJack
    Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack
    Wang, Hao, Li, Hanchen, Mang, Qiuyang, Cheung, Alvin, Sen, Koushik, and Song, Dawn
    2026
    ArXiv 𝕏
  4. Revelio
    Revelio: Cost-Efficient Agentic Memory Safety Vulnerability Detection For Repository-Scale Codebases
    Hou, Yiwei,  Wang, Hao, Lyu, Muxi, Momeu, Marius, Nguyen, Eric, Yang, Taige, Sen, Koushik, Song, Dawn, and Wagner, David
    2026
    ArXiv
  5. SCGAgent
    SCGAgent: Recreating the Benefits of Reasoning Models for Secure Code Generation with Agentic Workflows
    Saul, Rebecca,  Wang, Hao, Sen, Koushik, and Wagner, David
    2025
    ArXiv
  6. SecPI
    SecPI: Secure Code Generation with Reasoning Models via Security Reasoning Internalization
    Wang, Hao, Mündler, Niels, Vero, Mark, He, Jingxuan, Song, Dawn, and Vechev, Martin
    2026
    ArXiv
  7. Type
    Type-Constrained Code Generation with Language Models
    Mündler, Niels, He, Jingxuan,  Wang, Hao, Sen, Koushik, Song, Dawn, and Vechev, Martin
    In PLDI 2025
    ArXiv