Zheyuan "Brian" Zhang

I am currently doing a research intern on Embodied AI, supervised by Professor Chuang Gan. I graduated with an MS in Robotics from the Robotics Institute and a Graduate Certificate in Cognitive Science from the Weinberg Institute for Cognitive Science at the University of Michigan. I was a research assistant at the Situated Language and Embodied Dialogue (SLED) lab, advised by Professor Joyce Chai. I was also fortunate to work with Professor Honglak Lee, Professor Freda Shi, Professor Parisa Kordjamshidi. Prior to UMich, I did a BS in Computer Science from UMass Amherst, College of Information & Computer Sciences.

Email  /  CV  /  Scholar  /  Twitter  /  Github

profile photo

Research

My research is centered at the intersection of Artificial Intelligence and Cognitive Science. Currently, I am particularly interested in Embodied AI (reasoning, planning and acting in simulation and physical world, interactive learning, population, evolutionary computation), Deep Learning (algorithm and neural architecture for learning better representation, interpretability), Natural Language Processing (language grounding to vision and other modalities, evaluation of foundation language / multimodal models). I am also planning to explore AI for Science and AI for Psychology in the future.

News

  • [2024.10] COMFORT accepted to Pluralistic Alignment @ NeurIPS 2024.
  • [2024.10] EILeV accepted to Video-Language Models @ NeurIPS 2024.
  • [2024.09] EILeV accepted to EMNLP 2024.
  • [2023.10] HAR accepted to EMNLP 2023.

Preprints

img_comfort Do Vision-Language Models Represent Space and How? Evaluating Spatial Frame of Reference Under Ambiguities
Zheyuan Zhang*, Fengyuan Hu*, Jayjun Lee*, Freda Shi, Parisa Kordjamshidi, Joyce Chai, Ziqiao Ma
Pluralistic Alignment @ NeurIPS 2024
Project Page | Paper | Code | Dataset
img_combo A Combinatorial Approach to Neural Emergent Communication
Zheyuan Zhang
arXiv, 2024
Paper
img_combo COMBO: Compositional World Models for Embodied Multi-Agent Cooperation
Hongxin Zhang*, Zeyuan Wang*, Qiushi Lyu*, Zheyuan Zhang, Sunli Chen, Tianmin Shu, Yilun Du, Chuang Gan
arXiv, 2024
Project Page | Paper | Code

Publications

img_eilev EILeV: Eliciting In-Context Learning in Vision-Language Models for Videos Through Curated Data Distributional Properties
Keunwoo Peter Yu, Zheyuan Zhang, Fengyuan Hu, Shane Storks, Joyce Chai
Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024
Video-Language Models @ NeurIPS 2024
Paper | Code
img_har HAR: From Heuristic to Analytic: Cognitively Motivated Strategies for Coherent Physical Commonsense Reasoning
Zheyuan Zhang*, Shane Storks*, Fengyuan Hu, Sungryull Sohn, Moontae Lee, Honglak Lee, Joyce Chai
Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023
Paper | Code
img_lsvrg Low-cost Solution for Vision-based Robotic Grasping
Zheyuan Zhang, Huiliang Shang
International Conference on Networking Systems of AI (INSAI), IEEE Computer Society, 2021
(Second Prize)
Paper | Code (110 )

Other Projects

img_botlab Bot Lab: Autonomous Ground Vehicle from Low-level Control, SLAM to Planning and Exploration
Zheyuan Zhang, Yu Zhu, Manu Aatitya Raajan Priyadharshini, Thirumalaesh Ashokkumar
ROB 550 (Robotic Systems Laboratory), University of Michigan, 2022
Paper

Teaching

img_umich Graduate Student Instructor (GSI), EECS 492 (Introduction to Artificial Intelligence)
Winter 2024, Fall 2023

Textbook: Artificial Intelligence: A Modern Approach, 4th Edition

Academic Services

  • Conference Reviewer: ICLR 2025, NeurIPS 2024 Workshop

Misc

  • Placeholder

Source code borrowed from Jon Barron's website.