Publications

You can also find my articles on my Google Scholar profile

Research Outlines

My main research interests lie in the following aspects. Large Language models and foundation models. i. Model editing and memory management for large language models (LLMs). ii. LLM agent and reasoning. iii. Parametric understanding of LLMs (localization, merging, scaling, pruning, stitching, unlearning, editing, and etc.) iv. Text-to-model generation by diffusion transformers (DiTs). v. Vision-Language (V-L) representation and understanding of multi-modal foundation models. Trustworthy deep learning. i. Privacy-preserving federated learning (FL), efficient & robust algorithm design, and generalization, personalization & training dynamics understanding. ii. Mechanistic interpretability of neural networks, weight decay, loss landscape, permutation invariance, linear mode connectivity, and etc. iii. Socio-technical issues brought by collaborative learning. iv. Responsible and trustworthy AI.

Main Publications

* indicates equal contributions, # indicates corresponding author.

2025

• [18] Towards Universal Personalization in Federated Learning via Collaborative Foundation Generative Models
  Chenrui Wu*, Zexi Li*, Fangxin Wang, Hongyang Chen, Jiajun Bu, and Haishuai Wang#
  IEEE Transactions on Mobile Computing. (CCF A)

2024

• [17] Text-to-Model: Text-Conditioned Neural Network Diffusion for Train-Once-for-All Personalization
  Zexi Li*, Lingzhi Gao*, Chao Wu#
  preprint. [arxiv]
• [16] WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models
  Peng Wang*, Zexi Li*, Ningyu Zhang#, Ziwen Xu, Yunzhi Yao, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen#
  NeurIPS 2024. [arxiv]
• [15] Improving Group Connectivity for Generalization of Federated Deep Learning
  Zexi Li*, Jie Lin*, Zhiqi Li*, Didi Zhu, Tao Lin#, Chao Wu#
  preprint. [arxiv]
• [14] Model Tailor: Mitigating Catastrophic Forgetting in Multi-modal Large Language Models
  Didi Zhu, Zhongyi Sun, Zexi Li, Tao Shen, Ke Yan, Shouhong Ding, Kun Kuang#, Chao Wu#
  International Conference on Machine Learning (ICML) 2024. (CCF A, Top Conference in Machine Learning) [arxiv]
• [13] Neural Collapse Anchored Prompt Tuning for Generalizable Vision-Language Models
  Didi Zhu, Zexi Li, Min Zhang, Junkun Yuan, yunfeng shao, Yinchuan Li, Jiashuo Liu, Kun Kuang, Chao Wu#
  ACM KDD 2024. (CCF A, Top Conference in Data Mining) [arxiv]
• [12] OpenFedLLM: Training Large Language Models on Decentralized Private Data via Federated Learning
  Rui Ye, Wenhao Wang, Jingyi Chai, Dihan Li, Zexi Li, Yinda Xu, Yaxin Du, Yanfeng Wang, Siheng Chen
  ACM KDD 2024. (CCF A, Top Conference in Data Mining) [arxiv]

2023

• [11] FediOS: Decoupling Orthogonal Subspaces for Personalization in Feature-skew Federated Learning
  Lingzhi Gao*, Zexi Li*, Yang Lu, and Chao Wu#
  preprint. [arxiv]
• [10] Training-time Neuron Alignment through Permutation Subspace for Improving Linear Mode Connectivity and Model Fusion
  Zexi Li, Zhiqi Li, Jie Lin, Tao Shen, Tao Lin#, and Chao Wu#
  preprint. [arxiv]
• [9] Revisiting Weighted Aggregation in Federated Learning with Neural Networks
  Zexi Li, Tao Lin#, Xinyi Shang, and Chao Wu#
  International Conference on Machine Learning (ICML) 2023. (CCF A, Top Conference in Machine Learning) [paper][github][arxiv]
• [8] No Fear of Classifier Biases: Neural Collapse Inspired Federated Learning with Synthetic and Fixed Classifier
  Zexi Li, Xinyi Shang, Rui He, Tao Lin#, and Chao Wu#
  International Conference on Computer Vision (ICCV) 2023. (CCF A, Top Conference in Computer Vision) [paper][arxiv]
• [7] Universal Domain Adaptation via Compressive Attention Matching
  Didi Zhu*, Yinchuan Li*, Junkun Yuan, Zexi Li, Kun Kuang#, Chao Wu#
  International Conference on Computer Vision (ICCV) 2023. (CCF A, Top Conference in Computer Vision) [paper]
• [6] Edge-cloud Collaborative Learning with Federated and Centralized Features
  Zexi Li*, Qunwei Li*, Yi Zhou, Wenliang Zhong#, Guannan Zhang, and Chao Wu#
  International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) 2023. (CCF A, Top Conference in Data Mining and Information Retrieval) [paper][arxiv]
• [5] Learning Cautiously in Federated Learning with Noisy and Heterogeneous Clients
  Chenrui Wu*, Zexi Li*, Fangxin Wang#, and Chao Wu#
  (Oral) IEEE International Conference on Multimedia and Expo (ICME) 2023. (CCF B, Top Conference in Multimedia) [paper][arxiv] –

2022

• [4] Towards Effective Clustered Federated Learning: A Peer-to-peer Framework with Adaptive Neighbor Matching
  Zexi Li, Jiaxun Lu, Shuang Luo, Didi Zhu, Yunfeng Shao, Yinchuan Li, Zhimeng Zhang, Yongheng Wang#, and Chao Wu#
  • Long version in IEEE Transactions on Big Data. (JCR Q1, IF: 7.2) [paper][arxiv]
  • Short version in MLSys2022 Workshop. [poster]
• [3] Can We Share Models If Sharing Data Is Not an Option?
  Zexi Li, Feng Mao#, and Chao Wu#
  Patterns, Cell Press. (JCR Q1, IF: 6.5) [paper]

2021

• [2] Boosting the generalization ability of Vis-NIR-spectroscopy-based regression models through dimension reduction and transfer learning
  Xiaoli Li*, Zexi Li*, Xufeng Yang, and Yong He#
  Computers and Electronics in Agriculture. (JCR Q1, CAS Top, IF: 8.3) [paper]
• [1] An early assessment of the County Medical Community reform in China: a case study of Zhejiang province
  Chao Wu, Yixin Tu, Zexi Li, Jianxing Yu#
  Journal of Chinese Governance. (SSCI, JCR Q1, IF: 3.0) [paper]