Hi there! I’m an incoming PhD student at the University of California, Santa Barbara, Department of Computer Science, advised by Dr. Eric Xin Wang. I am also in my final year as an undergraduate student at the School of Engineering, University of Liverpool.
My research interests lie in trustworthy AI and the algorithmic foundations of multimodal large models, with a focus on enhancing their reasoning capabilities, reliability, and interpretability across diverse input modalities. I am always open to collaboration and the exchange of ideas. If you’d like to discuss potential research opportunities or simply connect, please don’t hesitate to reach out to me at chengzhi@ucsb.edu.
🔥 News
- 2025.5 🎉🎉 We release a paper on balancing reasoning and hallucination in multimodal reasoning models. 👉 Check our paper and project
- 2025.4 🎉🎉 One papes is accepted by ACL 2025 , thanks for all of my collaborators. Check it out in the paper.
- 2025.4 🎉🎉 One papes is accepted by CVPR 2025 , thanks for all of my collaborators. Check it out in the paper.
- 2025.2 🎉🎉 Two papers are accepted by ICLR 2025, thanks for all of my collaborators. Check out our papers MSSBench and ANTRP.
- 2025.1 🎉🎉 One paper is accepted by AAAI 2025 Oral, thanks for all of my collaborators. Check it out in the paper.
💥 Preprints
More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Models
Chengzhi Liu*, Zhongxing Xu*, Qinyue Wei, Juncheng Wu, James Zou, Xin Eric Wang, Yuyin Zhou, Sheng Liu
Project |
The hidden risks of large reasoning models: A safety assessment of r1
Kaiwen Zhou*, Chengzhi Liu*, Xuandong Zhao, Shreedhar Jangam, Jayanth Srinivasa, Gaowen Liu, Dawn Song, Xin Eric Wang
📝 Selected Publications
Multimodal Situational Safety
Kaiwen Zhou*, Chengzhi Liu*, Xuandong Zhao, Anderson Compalas, Dawn Song, Xin Eric Wang
Project |
Incomplete Modality Disentangled Representation for Ophthalmic Disease Grading and Diagnosis
Chengzhi Liu*, Zile Huang*, Zhe Chen, Feilong Tang, Yu Tian, Zhongxing Xu, Zihong Luo, Yalin Zheng, Yanda Meng
Intervening anchor token: Decoding strategy in alleviating hallucinations for MLLMs
Feilong Tang*, Zile Huang*, Chengzhi Liu, Qiang Sun, Harry Yang, Ser-Nam Lim
Project |
📖 Educations
- 2021.09 - 2025.07, Bachelor, University of Liverpool, Liverpool, United Kingdom.
- 2025.09 - 2030.06, Ph.D., University of California, Santa Barbara, Santa Barbara, CA, United States.