Kaiya Ivy Zhao

Kaiya Ivy Zhao

EECS PhD Student @ MIT

Massachusetts Institute of Technology

kyzhao@mit.edu

Biography

Hi! I’m Ivy, a PhD student at MIT CSAIL where I am fortunate to be co-advised by Phillip Isola and Josh Tenenbaum. Previously, I had a wonderful experience working with Guangyu Robert Yang at MIT on autonomous agents with real-time interaction.

My research interest lies in the intersection of AI and human intelligence, with a focus on AI systems that can understand and interact with humans in a natural manner. I am passionate about generalist agents in physical and social settings.

Specifically, I am curious about the following topics:

  • Embodied Intelligence and World Representations: What is the intuitive physical engine in our brain that helps understand the physical world? How can we endow AI agents with world model that simulates the environment and the ability to perceive, reason and act in the physical world?
  • Cognitive Science and Social Reasoning: What are the underlying mechanisms to infer human intention (Theory-of-Mind)? How can we leverage insights from human cognition to design human-like agents and improve collaboration?

I publish under the name “Kaiya Ivy Zhao”, where “Zhao” is my surname and “Kaiya” is the forename. “Ivy”(ai·vee) is the name I usually go by.

Outside the research, my world is filled with the vibrancy of musicals, the companionship of cats and the thrill of travel. Each of these hobbies offers me a unique perspective on life and creativity.

Feel free to reach out via email to discuss research, collaborations, or schedule a chat!

Interests
  • Embodied Intelligence
  • Physical Reasoning
  • Cognitive Science
  • Social Reasoning
Education
  • PhD in Computer Science, 2024 - present

    Massachusetts Institute of Technology

  • BSc in Computer Science, 2020 - 2024

    Fudan University

Recent Publications

Quickly discover relevant content by filtering publications.
(2025). Assessing Adaptive World Models in Machines with Novel Games.

Cite arXiv

(2025). Cross-Modal Alignment Regularization: Enhancing Language Models with Vision Model Representations. Second Workshop on Representational Alignment @ICLR 2025.

Cite URL

(2023). Enhancing Understanding in Generative Agents through Active Inquiring. Intrinsically-Motivated and Open-Ended Learning Workshop @NeurIPS2023.

Cite URL

Contact

Feel free to get in touch with me using the contact options below.