Hangoo Kang

About Me

I’m Hangoo Kang, a first-year M.S. student in Computer Science (MSCS) at Stanford University.

I work in the Scaling Intelligence Lab with Professor Azalia Mirhoseini, focusing on continuous learning for agentic AI and AI reasoning. I’m also part of xLab, advised by Professor Yejin Choi, where I study model collapse and trustworthiness in multi-turn environments. Before Stanford, I worked with Professor Gagandeep Singh, Professor Sasa Misailovic, and Professor Minjia Zhang on safety in agentic AI systems, grammar-constrained decoding, and mitigating reward hacking in RLHF.

Research Interests

• Machine Learning: Trustworthy AI, Agents, Multimodal Models, Post-training
• Programming Languages: Grammar-Guided Generation, Software Verification, Program Synthesis
• Reinforcement Learning: Continuous Learning, RLHF, Robust Verifiers

News

• [Sept. 2025] Our paper “Learning a Pessimistic Reward Model in RLHF” was accepted to ALERT@NeurIPS.
• [Sept. 2025] Our paper TRAP was accepted to NeurIPS 2025.
• [Aug. 2025] Started my M.S. at Stanford University.
• [May 2023] Started as an undergraduate RA at UIUC’s ARC Lab.
• [May 2023] Started as an undergraduate RA at UIUC’s FOCAL Lab.

Publications

Powered by Jekyll and Minimal Light theme.