First-Year PhD · NExT++ Lab, National University of Singapore

Chenhang Cui

Probing latent knowledge inside large foundation models and building capable agents on top of them.

I am a first-year PhD student at the NExT++ Research Centre, School of Computing, NUS, advised by Prof. Chua Tat-Seng, working closely with Dr. Fei Shen.

Before starting my PhD, I interned at NExT++ with Dr. An Zhang, and worked as a research intern with Prof. Huaxiu Yao at UNC. I received my B.Eng. (Honors) from UESTC, where I was advised by Prof. Yazhou Ren.

My current research centers on latent knowledge & interpretability of large foundation models (neuron-level mechanisms, cross-modal transfer) and autonomous LLM/VLM agents. I am also broadly interested in trustworthy LFMs and multi-view representation learning.

Curriculum Vitae Google Scholar GitHub Email
Chenhang Cui
  • Singapore
  • NUS · NExT++ Lab
  • Trustworthy VLMs · RL Alignment
  • 中文 / English

# Research Interests

Agents

Autonomous LLM/VLM agents — tool use, reasoning, planning, and multi-agent collaboration.

Latent Knowledge

Interpretability of large models: neuron-level mechanisms, cross-modal transfer, and internal representations.

Trustworthy LFMs

Hallucination, jailbreak, safety alignment, and robustness of large foundation models.

Representation Learning

Multi-view & multimodal representation learning — consistency, complementarity, and cross-architecture transfer.

# News

# Selected Publications

* denotes equal contribution. Full list on Google Scholar.

First-author & Co-first-author

CVPR-F
2026

Do LLMs and VLMs Share Neurons for Inference? Evidence and Mechanisms of Cross-Modal Transfer

Chenhang Cui, An Zhang, Yuxin Chen, Gelei Deng, Jingnan Zheng, Zhenkai Liang, Xiang Wang, Tat-Seng Chua

ICML
2026

Transport and Merge: Cross-Architecture Merging for Large Language Models

Chenhang Cui, Binyun Yang, Fei Shen, Yuxin Chen, Jingnan Zheng, Xiang Wang, An Zhang, Tat-Seng Chua

NeurIPS
2025

Safe + Safe = Unsafe? Exploring How Safe Images Can Be Exploited to Jailbreak Large Vision-Language Models

Chenhang Cui, Gelei Deng, An Zhang, Yicong Li, Lianli Gao, Tianwei Zhang, Tat-Seng Chua

arXiv
2025

Fading Focus: Mitigating Visual Attention Degradation in Large Vision-Language Models

Chenhang Cui, Jiabing Yang, Yiyang Zhou, Peng Xia, Ying Wei, Huaxiu Yao

ICLR
2025

Fine-Grained Verifiers: Preference Modeling as Next-Token Prediction in Vision-Language Alignment

Chenhang Cui, An Zhang, Yiyang Zhou, Zhaorun Chen, Gelei Deng, Huaxiu Yao, Tat-Seng Chua

ECCV
2024

How Many Are in This Image? A Safety Evaluation Benchmark for Vision LLMs

Haoqin Tu*, Chenhang Cui*, Zijun Wang*, Yiyang Zhou, Bingchen Zhao, Junlin Han, Wangchunshu Zhou, Huaxiu Yao, Cihang Xie

ICLR
2024

Analyzing and Mitigating Object Hallucination in Large Vision-Language Models

Yiyang Zhou*, Chenhang Cui*, Jaehong Yoon, Linjun Zhang, Zhun Deng, Chelsea Finn, Mohit Bansal, Huaxiu Yao

ICLR-W
2024

Aligning Modalities in Vision Large Language Models via Preference Fine-tuning

Yiyang Zhou*, Chenhang Cui*, Rafael Rafailov, Chelsea Finn, Huaxiu Yao

NeurIPS
2023

A Novel Approach for Effective Multi-View Clustering with Information-Theoretic Perspective

Chenhang Cui, Yazhou Ren, Jingyu Pu, Jiawei Li, Xiaorong Pu, Tianyi Wu, Yutao Shi, Lifang He

IJCAI
2023

Deep Multi-view Subspace Clustering with Anchor Graph

Chenhang Cui, Yazhou Ren, Jingyu Pu, Xiaorong Pu, Lifang He

arXiv
2023

Holistic Analysis of Hallucination in Large Vision-Language Models: Bias and Interference Challenges

Chenhang Cui*, Yiyang Zhou*, Xiangyu Yang, Shirley Wu, Linjun Zhang, James Zou, Huaxiu Yao

Co-author

TPAMI
2026

Enhancing Multi-View Clustering: A Sufficient Information-Theoretic Approach

Yazhou Ren, Zichen Wen, Junlong Ke, Chenhang Cui, Yonghao Huang, Xinyue Chen, Philip S. Yu, Lifang He

NeurIPS
2026

RSafe: Incentivizing Proactive Reasoning to Build Robust and Adaptive LLM Safeguards

Xiangtian Ji, Yijun Lu, Chenhang Cui, Weixiang Zhao, Gelei Deng, Zhenkai Liang, An Zhang, Tat-Seng Chua

ICCV
2025

VFlowOpt: A Token Pruning Framework for LMMs with Visual Information Flow-Guided Optimization

Sihan Yang, Runsen Xu, Chenhang Cui, Tai Wang, Dahua Lin, Jiangmiao Pang

EMNLP-F
2025

Improving Alignment in LVLMs with Debiased Self-Judgment

Sihan Yang, Chenhang Cui, Zihao Zhao, Yiyang Zhou, Weilong Yan, Ying Wei, Huaxiu Yao

TMLR
2025

Reliable and Responsible Foundation Models

Xinyu Yang, Junlin Han, Rishi Bommasani, Jinqi Luo, Wenjie Qu, Wangchunshu Zhou, Adel Bibi, Xiyao Wang, Jaehong Yoon, ..., Chenhang Cui, et al., Huaxiu Yao

ICLR
2025

MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models

Peng Xia, Siwei Han, Shi Qiu, Yiyang Zhou, Zhaoyang Wang, Wenhao Zheng, Zhaorun Chen, Chenhang Cui, Mingyu Ding, Linjie Li, Lijuan Wang, Huaxiu Yao

NeurIPS
2024

Calibrated Self-Rewarding Vision Language Models

Yiyang Zhou, Zhiyuan Fan, Dongjie Cheng, Sihan Yang, Zhaorun Chen, Chenhang Cui, Xiyao Wang, Yun Li, Linjun Zhang, Huaxiu Yao

AAAI
2024

Adaptive Feature Imputation with Latent Graph for Deep Incomplete Multi-View Clustering

Jingyu Pu, Chenhang Cui, Xinyue Chen, Yazhou Ren, Xiaorong Pu, Zhifeng Hao, S. Yu Philip, Lifang He

ICML-W
2024

MJ-Bench: Is Your Multimodal Reward Model Really a Good Judge?

Zhaorun Chen, Yichao Du, Zichen Wen, Yiyang Zhou, Chenhang Cui, et al.

IJCAI
2024

Dynamic Weighted Graph Fusion for Deep Multi-View Clustering

Yazhou Ren, Jingyu Pu, Chenhang Cui, Yan Zheng, Xinyue Chen, Xiaorong Pu, Lifang He

IJCAI
2024

Integrating Vision-Language Semantic Graphs in Multi-View Clustering

Junlong Ke, Zichen Wen, Yechenhao Yang, Chenhang Cui, Yazhou Ren, Xiaorong Pu, Lifang He

ACM MM
2024

Dual-Optimized Adaptive Graph Reconstruction for Multi-View Graph Clustering

Zichen Wen, Tianyi Wu, Yazhou Ren, Yawen Ling, Chenhang Cui, Xiaorong Pu, Lifang He

# Research Experience

PhD · Latent Knowledge & Agents

NExT++ Lab, NUS · Jan 2026 – Present · Advisor: Prof. Chua Tat-Seng · with Dr. Fei Shen

Interpretability and latent-knowledge mechanisms of large foundation models; autonomous LLM/VLM agents.

Research Intern · Safety & Multimodal RL Alignment

NExT++ Lab, NUS · Aug 2024 – Aug 2025 · Advisors: Prof. Chua Tat-Seng, Prof. Zhenkai Liang · Host: Prof. An Zhang

Enhancing safety and alignment of VLLMs; outputs accepted to ICLR 2025 and NeurIPS 2025.

Research Intern · Hallucination in Large VLMs

UNC MURGe-Lab · May 2023 – Jan 2024 · Host: Prof. Huaxiu Yao

Reduced object hallucination by 23% on MiniGPT-4 without sacrificing diversity; ICLR 2024 + NeurIPS 2023 Workshop.

Undergraduate Research · Multi-view Representation Learning

UESTC · May 2022 – Sep 2023 · Advisor: Prof. Yazhou Ren

Consistency & complementarity in multimodal representation; three first-author papers (NeurIPS / IJCAI / AAAI).

# Honors & Services

Honors & Awards

  • SenseTime Scholarship — top 25 nationwide, 2024
  • HUAWEI Scholarship, 2023
  • UESTC Academic Scholarship, 2022 / 2023 / 2024
  • MCM Honorable Mention (top 15%), Student Advisor, 2024

Reviewer Service

  • ICLR 2025 / 2026, CVPR 2025
  • KDD 2024, ARR 2024
  • Neural Networks (journal)