A one-sentence research thesis

I build and evaluate AI agents for cybersecurity.

Hwiwon Lee

PhD Student in Computer Science at the University of Illinois Urbana-Champaign · advised by Lingming Zhang.

Portrait of Hwiwon Lee

01Research

Systems that can reason, act, and be measured.

  1. Agentic Security

    Building and evaluating autonomous agents for realistic, long-horizon security engineering workflows.

  2. Software Systems

    Designing reproducible systems that connect models, program analysis, and dynamic execution evidence.

  3. Applied Vulnerability Research

    Discovering and validating vulnerabilities in production software, from services to operating-system kernels.

02Selected work

Recent publications.

  1. arXiv · 2026

    SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?

    Hwiwon Lee, Jiawei Liu, Dongjun Kim, Wubing Xia, Ziqi Zhang, Chunqiu Steven Xia, Lingming Zhang

    PDF: SEC-bench Pro

  2. arXiv · 2026

    Agentic Vulnerability Reasoning on COTS Binaries

    Hwiwon Lee, Jongseong Kim, Lingming Zhang

    PDF: Agentic Vulnerability Reasoning on COTS Binaries

  3. NeurIPS · 2025

    SEC-bench: Automated Benchmarking of LLM Agents on Real-World Software Security Tasks

    Hwiwon Lee, Ziqi Zhang, Hanxiao Lu, Lingming Zhang

    PDF: SEC-bench

  4. IEEE S&P · 2024 · Distinguished Paper

    BENZENE: A Practical Root Cause Analysis System with an Under-Constrained State Mutation

    Younggi Park, Hwiwon Lee, Jinho Jung, Kevin Koo, Huy Kang · Distinguished Paper

    PDF: BENZENE

03News

What’s new.

  1. SEC-bench Pro, our benchmark for long-horizon software security tasks, is now available on arXiv.

  2. Agentic Vulnerability Reasoning on COTS Binaries is now available on arXiv.

  3. Recognized as a Microsoft Most Valuable Security Researcher for 2026 Q1.