I’m Guanyu Zhou, a research intern at Princeton University and Falcon Vision Team at Technology Innovation Institute, currently a final year undergraduate student.

I am working closely with Prof. Zhuang Liu at Princeton University and Prof. Harry Yang at HKUST. My research interests are Multimodal Foundation Models (Visual Perception, Reasoning and AIGC). Prior to this, I did a research internship at the NLPGroup of HKUST, exploring the perception bottleneck of MLLMs/VLMs. I also had an unforgettable research time at AIMlab, where I worked on problems related to video understanding.

My ultimate goal is to build a multimodal, scalable, dataset-free, continuously learning, and self-iterating intelligent system. Feel free to email me to discuss issues in related fields!😄

📅 News

• [2026.06] Excited to be invited to join the Falcon Vision Team at Technology Innovation Institute (UAE TII) as a research intern！
• [2026.04] We released VisionFoundry!
• [2025.11] OmniDPO accepted by AAAI 2026.
• [2025.03] I made a sharing at AI TIME, welcome to watch!
• [2025.01] Our paper CausalMM has been accepted to ICLR 2025, see you in Singapore!
• [2024.07] I am visiting HKUST(GZ) under the supervision of Professor Xuming Hu!

📝 Publications

[ICLR 2025] Mitigating Modality Prior-Induced Hallucinations in Multimodal Large Language Models via Deciphering Attention Causality

Guanyu Zhou, Yibo Yan, Xin Zou, Kun Wang, Aiwei Liu, Xuming Hu

PDF Code Talk ICLR

VisionFoundry: Teaching VLMs Visual Perception with Synthetic Images

Guanyu Zhou, Yida Yin, Wenhao Chai, Shengbang Tong, Xingyu Fu, Zhuang Liu

Page PDF Code Dataset Blog

[AAAI 2026] OmniDPO: A Preference Optimization Framework to Address Omni-Modal Hallucination

Junzhe Chen, Tianshu Zhang, Shiyu Huang, Yuwei Niu, Chao Sun, Rongzhou Zhang, Guanyu Zhou, Lijie Wen, Xuming Hu

PDF Code

JLT: Clean-Latent Prediction in Latent Diffusion Transformers

Funing Fu, Tenghui Wang, Guanyu Zhou, Junyong Cen, Qichao Zhu

PDF Code Page HF Blog

OccludeNet: A Causal Journey into Mixed-View Actor-Centric Video Action Recognition under Occlusions

Guanyu Zhou, Wenxuan Liu, Wenxin Huang, Xuemei Jia, Xian Zhong, Chia-Wen Lin

PDF Code

CML-Bench: A Framework for Evaluating and Enhancing LLM-Powered Movie Scripts Generation

Mingzhe Zheng, Dingjie Song, Guanyu Zhou, Jun You, Jiahao Zhan, Xuran Ma, Xinyuan Song, Ser-Nam Lim, Qifeng Chen, Harry Yang

PDF Code Dataset

An Efficient Manifold-Hypothesis-Based Dimensional Analysis Algorithm and Its Applications in MLLMs

Guanyu Zhou, Yonghua Hei, Yibo Yan, Xin Zou, Junzhe Chen, Xuming Hu

Preparing

📇 Experience

[2025.4] Princeton University / ZLab

Topic: Synthetic Data and VLMs

[2024.7] HKUST / AI Thrust

Topic: Bias in Model Prior Knowledge

[2023.7] AIMLab@WHUT / Action Recognition Group

Topic: Bias in the Data (Learning Process)

✍ Services

Reviewer

• IEEE International Conference on Computer Vision 2025 (ICCV 2025)
• IEEE International Conference on Multimedia & Expo 2025 (ICME 2025)
• IEEE Transactions on Circuits and Systems for Video Technology (TCSVT)

Management

• Deputy Director of Competition Management Center of the School’s Innovation and Entrepreneurship Department (2023-2024)

👨‍🏫 Talks

• AI TIME Online Pre-Session (ICLR 2025) [watch on Bilibili]