Marx Wang

Hi! 👋 I'm a third year PhD student at the University of Washington Information School, advised by Alexis Hiniker and Anind Dey at the User Empowerment Lab.

My research sits at the intersection of HCI and LLM Affective Alignment. I study how people — especially young adults — relate to conversational AI, and what happens when those systems get emotions wrong. My work spans both mix-method empirical studies of how users engage with conversational agents and technical methods for auditing and modulating the emotional behavior of language models.

Some of my recent methods: qualitative conversation analysis grounded in developmental psychology, computational psycholinguistics/NLP (LIWC, NRC-VAD, RoBERTa), and mechanistic interpretability using Sparse Autoencoders to identify and steer affect-aligned features in LLM activations.

I also run ForYouPage.org, a youth-led nonprofit for mental health advocacy reaching 30,000+ members across 6 continents. I host conferences at UNGA🌍

Outside of research, you can find me doing MMA 🥊 and standup comedy (which collectively prepare me for paper rebuttals one day).

If any of this resonates — let's grab coffee ☕ and chat!