Bangzheng Li

Email: bzhli[at]ucdavis[dot]edu

👋 Hi, I am Bangzheng Li, a final-year PhD candidate in Computer Science at UC Davis (previously at Uniersity of Southern California). I'm a member of LUKA Lab where I am very grateful to be advised by Dr. Muhao Chen and funded by Provost's Fellowship.

My research focuses on Multimodal Large Language Models (MLLMs), with a particular emphasis on understanding and reasoning across text and vision. I am especially interested in the emergence of unified reasoning—where information flows seamlessly across all modalities. I believe that in the next generation of AI systems, the notion of “multimodality” will dissolve, giving rise to models that reason over any signal in a truly integrated manner.

I received my Bachelor's Degree from University of Illinois at Urbana-Champaign where I was fortunate to work with Dr. Jiawei Han on various text classification and document classification tasks.

I am graduating in Summer 2026 and am actively seeking full-time oppotunities. Open to work

📑 Selected Research Projects

Latent Visual Reasoning

Bangzheng Li, Ximeng Sun , Jiang Liu, Ze Wang, Jialian Wu, Xiaodong Yu, Hao Chen, Emad Barsoum, Muhao Chen, Zicheng Liu

preprint, 2025

[paper] [website] [code] [ model: LVR-7B]

QLIP: A Dynamic Quadtree Vision Prior Enhances MLLM Performance Without Retraining

Kyle R. Chickering, Bangzheng Li, Muhao Chen

preprint, 2025

[paper]

BLINK: Multimodal Large Language Models Can See but Not Perceive

Xingyu Fu*, Yushi Hu*, Bangzheng Li, Yu Feng, Haoyu Wang, Xudong Lin, Dan Roth, Noah A. Smith, Wei-Chiu Ma†, Ranjay Krishna†

ECCV 2024, Spotlight of cVinW@CVPR 2024, 36K total downloads.

[paper] [website] [code] [dataset] [eval] [twitter] [ Paper of the day]

Semantic-Clipping: Efficient Vision-Language Modeling with Semantic-Guidedd Visual Selection

Bangzheng Li, Fei Wang, Ben Zhou, Nan Xu, Sheng Zhang, Hoifung Poon, Muhao Chen

preprint, 2024

[paper]

Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?

Bangzheng Li, Ben Zhou, Fei Wang, Xingyu Fu, Dan Roth, Muhao Chen

NAACL. 2024.

[paper] [website] [code] [dataset]

Ultra-fine Entity Typing with Indirect Supervision from Natural Language Inference

Bangzheng Li,Wenpeng Yin, Muhao Chen

TACL. 2022.

[paper] [code]

FAMICOM: Further Demystifying Prompts for Language Models with Task-Agnostic Performance Estimation

Bangzheng Li, Bangzheng Li, Ben Zhou, Xingyu Fu, Fei Wang, Dan Roth, Muhao Chen

Arxiv 2024 June

[paper]

Affective and Dynamic Beam Search for Story Generation

Tenghao Huang, Ehsan Qasemi, Bangzheng Li, He Wang, Faeze Brahman, Muhao Chen, Snigdha Chaturvedi

EMNLP. 2023.

[paper] [code]

Unified semantic typing with meaningful label inference

James Y. Huang, Bangzheng Li, Jiashu Xu, Muhao Chen

NAACL. 2022.

[paper] [code]

Does Your Model Classify Entities Reasonably? Diagnosing and Mitigating Spurious Correlations in Entity Typing

Nan Xu, Fei Wang, Bangzheng Li, Mingtao Dong, Muhao Chen

EMNLP. 2022.

[paper]

COVID-19 Literature Knowledge Graph Construction and Drug Repurposing Report Generation

Qingyun Wang, Manling Li, Xuan Wang, Nikolaus Parulian, Guangxing Han, Jiawei Ma, Jingxuan Tu, Ying Lin, Ranran Haoran Zhang, Weili Liu, Aabhas Chauhan, Yingjun Guan, Bangzheng Li, Ruisong Li, Xiangchen Song, Yi Fung, Heng Ji, Jiawei Han, Shih-Fu Chang, James Pustejovsky, Jasmine Rah, David Liem, Ahmed ELsayed, Martha Palmer, Clare Voss, Cynthia Schneider, Boyan Onyshkevych

NAACL. 2021, Best Demo Paper

[paper] [website]

Neural Concept Map Generation for Effective Document Classification with Interpretable Structured Summarization

Carl Yang, Jieyu Zhang, Haonan Wang, Bangzheng Li, Jiawei Han

SIGIR. 2020.

[paper]
Comprehensive Named Entity Recognition on CORD-19 with Distant or Weak Supervision

Xuan Wang, Xiangchen Song, Bangzheng Li, Yingjun Guan, Jiawei Han

ISMB. 2020.

[paper]
Fine-Grained Named Entity Recognized Dataset of COVID-19 Literature

Xuan Wang, Xiangchen Song, Bangzheng Li, Kang Zhou, Qi Li, Jiawei Han

BIBM. 2020.

[paper]

Bangzheng Li

💼 Work Experience

📑 Selected Research Projects

Latent Visual Reasoning

QLIP: A Dynamic Quadtree Vision Prior Enhances MLLM Performance Without Retraining

BLINK: Multimodal Large Language Models Can See but Not Perceive

Semantic-Clipping: Efficient Vision-Language Modeling with Semantic-Guidedd Visual Selection

Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?

Ultra-fine Entity Typing with Indirect Supervision from Natural Language Inference

FAMICOM: Further Demystifying Prompts for Language Models with Task-Agnostic Performance Estimation

Affective and Dynamic Beam Search for Story Generation

Unified semantic typing with meaningful label inference

Does Your Model Classify Entities Reasonably? Diagnosing and Mitigating Spurious Correlations in Entity Typing

COVID-19 Literature Knowledge Graph Construction and Drug Repurposing Report Generation

Neural Concept Map Generation for Effective Document Classification with Interpretable Structured Summarization

Comprehensive Named Entity Recognition on CORD-19 with Distant or Weak Supervision

Fine-Grained Named Entity Recognized Dataset of COVID-19 Literature