Yian Wang

Yian Wang

PhD candidate · UIUC Computer Science · advised by Hari Sundaram and Varun Chandrasekaran

01About

I am a PhD candidate in Computer Science at the University of Illinois Urbana-Champaign, advised by Prof. Hari Sundaram and Prof. Varun Chandrasekaran. I started my PhD in Fall 2023.

My research lies in trustworthy machine learning and AI alignment I study how learning systems encode, retain, and propagate concepts or undesirable behaviors, and how these mechanisms can be controlled through unlearning and causal analysis.

I am currently interested in machine unlearning, concept representations in generative models, and emergent behavior in multi-agent systems, especially settings where harmful behavior or hidden influence can persist, transfer, or emerge through interaction.

Before UIUC, I earned my B.S. in Physics from the University of Science and Technology of China (USTC).

02Research Interests

  • Trustworthy Machine Learning
  • Machine Unlearning
  • Concept Representations
  • LLM alignment
  • Multi-Agent Systems
  • Human-AI Collaboration

03News

  • Our paper “Unlearning Isn’t Forgetting: Revealing Hidden Leakage in Class Unlearning Evaluations” was accepted to ICML 2026.
  • Our paper “From Plausible to Causal: Counterfactual Semantics for Policy Evaluation in Simulated Online Communities” received a Best Paper Award at the PoliSim Workshop @ CHI 2026.
  • Our paper “CausalDetox: Causal Head Selection and Intervention for Language Model Detoxification” was accepted to findings of ACL 2026.
  • Our paper on strategic antisocial behavior online was accepted to CSCW 2025.
  • Started my PhD at UIUC.

04Publications

05Contact

email scholar github linkedin cv