Zhihao Chen
I am an MSc student in Data Science at EPFL and a Graduate Student Researcher at VILAB. My research centers on multimodal foundation models and representation learning. While LLMs are powerful, modeling vision purely through language-centric approaches feels counter-intuitive. I believe multimodal intelligence requires efficient representations that genuinely leverage rich perceptual signals. I am interested in how foundation models can learn shared representations across modalities to reason and generate in a unified way, making them more adaptable, reliable, and useful in open-ended real-world settings.
I earned a BSc in Artificial Intelligence from the Polytechnic University of Catalonia (UPC). Previously, I completed my bachelor thesis as a Research Intern at CVC-UAB, under the supervision of Joost van de Weijer.
Outside of research, I enjoy working out, hiking, and drawing ;).
I am currently looking for research intern opportunities starting in Summer 2027. Feel free to reach out if you’d like to chat!
news
| Sep 30, 2026 | Our paper, “Continual Adaptation Across Modalities: Learning and Forgetting in Unified Multimodal Models”, was accepted to the NeurIPS 2026 Workshop on CL4FMAgents. The preprint is available on arXiv:2610.38436. |
|---|---|
| Sep 15, 2026 | Joined VILAB as a Graduate Student Researcher, working on unified multimodal modelling. |
| Sep 07, 2026 | Started the MSc in Data Science at EPFL. |
| Sep 07, 2026 | Started as a Student Teaching Assistant for EE-556 Mathematics of Data: From Theory to Computation at EPFL. |
| Jul 01, 2026 | Defended my bachelor thesis and graduated with a BSc in Artificial Intelligence from UPC. |