About Me
Welcome to Chenhui Xu’s home page.
I am now a doctoral student in University of Maryland, College Park. Before that, I was a PhD student at X-Lab of Department of Computer Science and Engineering at Univeristy at Buffalo. Prior to this, I spent two years in IF Lab of Department of Electrical and Computer Engineering at Geroge Mason University under the guidance of Prof. Xiang Chen. I received my Bachelor’s degree in Statistics from University of Science and Technology of China.
I am advised by Prof. Heng Huang. I was advised by Prof. Jinjun Xiong at University at Buffalo and co-advised by Prof. Xiang Chen at Peking University.
My current research interests include large multimodal language model post-training, especially on reinforcement learning, on-policy distillation. I have experience training models with large clusters (500+ GPUs). I foucus on applying advanced machine learning techniques to multiple applications, including geospatial AI, physical AI and recommendations, while keeping outputs innovations to the machine learning community.
News
- 04/2026: One paper on Geospatial Reasoning Vision Language Model is accepted by ICML 2026..
- 02/2026: I will move to University of Maryland to continue my Ph.D in Fall 2026.
- 01/2026: One collaborative paper is accepted by ICLR 2026.
- 09/2025: One paper on Physics-Informed Neural Networks is accepted by NeurIPS 2025.
- 05/2025: I will work as a research intern in Microsoft during summer 2025.
- 05/2025: Check our new preprint paper discussing the impact of arithmetic precision on Physics-Informed Neural Networks.
- 05/2025: One paper on Physics-Informed Neural Networks is accepted by ICML 2025.
...
- 04/2025: One collaborative paper is accepted by IJCAI 2025 AI and Social Good Track.
- 02/2025: One collaborative paper is accepted by DAC 2025.
- 01/2025: One collaborative paper is accepted by ICLR 2025.
- 09/2024: One paper is accepted by NeurIPS 2024.
- 05/2024: One paper on Out-of-Distribution Detection is accepted by ICML 2024.
- 01/2024: I will join University at Buffalo to continue my Ph.D. career under the guidance of Prof. Jinjun Xiong.
- 10/2023: One paper is accepted by ASP-DAC 2024 and received Best Paper Nomination.
- 10/2022: Serve as session chair of IEEE Cloud Summit.
- 07/2022: I left USTC and will join George Mason University (GMU) to pursue a Ph.D. degree under guidance of Dr. Xiang Chen.
- 06/2021: I graduated and obtained my bachelor’s degree from University of Science and Technology of China.
Publications
- [ICML 2026] Chenhui Xu, Fuxun Yu, Michael J Bianco, Jacob Kovarskiy, Raphael Tang, Qi Zhang, Zirui Xu, Will LeVine, Brandon Dubbs, Heming Liao, Cassandra Burgess, Suvam Bag, Jay Patravali, Rupanjali Kukal, Mikael Figueroa, Rishi Madhok, Nikolaos Karianakis, Jinjun Xiong. Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewards, in International Conference on Machine Learning, 2026.
- [ICLR 2026] Yuting Hu, Tarek Mohamed, Chenhui Xu, Hua Xiang, Hussam Amrouch, Gi-Joon Nam, Jinjun Xiong. Si-GT: Fast Interconnect Signal Integrity Analysis for Integrated Circuit Design via Graph Transformers, in International Conference on Learning Representations, 2026.
- [WACV 2026] Chenhui Xu, Fuxun Yu, Jinjun Xiong, Xiang Chen. QuadraNet V2: Efficient and Sustainable Training of High-Order Neural Networks with Quadratic Adaptation, in Winter Conference on Applications of Computer Vision, 2026.
- [NeurIPS 2025] Chenhui Xu, Dancheng Liu, Amir Nassereldine, Jinjun Xiong. FP64 is All You Need: Rethinking Failure Modes in Physics-Informed Neural Networks, in Annual Conference on Neural Information Processing Systems, 2025.
- [ICCAD 2025] Ruiyang Qin, Dancheng Liu, Gelei Xu, Zheyu Yan, Chenhui Xu, Yuting Hu, X Sharon Hu, Jinjun Xiong, Yiyu Shi. Tiny-Align: Bridging Automatic Speech Recognition and Large Language Model on the Edge, IEEE/ACM International Conference on Computer-Aided Design, 2025.
- [ICML 2025] Chenhui Xu, Dancheng Liu, Yuting Hu, Jiajie Li, Ruiyang Qin, Qingxiao Zheng, Jinjun Xiong. Sub-Sequential Physics-Informed Learning with State Space Model, in International Conference on Machine Learning, 2025.
- [IJCAI 2025] Yuting Hu, Dancheng Liu, Qingyun Wang, Charles Yu, Chenhui Xu, Qingxiao Zheng, Heng Ji, Jinjun Xiong. Automating Intervention Discovery from Scientific Literature: A Progressive Ontology Prompting and Dual-LLM Framework, in International Joint Conference on Artificial Intelligence, AI and Social Good Track, 2025.
- [DAC 2025] Dancheng Liu, Chenhui Xu, Jiajie Li, Amir Nassereldine, Jinjun Xiong. Ensembler: Protect Collaborative Inference Privacy from Model Inversion Attack via Selective Ensemble, in Design Automation Conference, 2025.
- [ICLR 2025] Jiajie Li, Brian R Quaranto, Chenhui Xu, Ishan Mishra, Ruiyang Qin, Dancheng Liu, Peter C W Kim, Jinjun Xiong. Recognize Any Surgical Object: Unleashing the Power of Weakly-Supervised Data, in International Conference on Learning Representations, 2025. (Spotlight. 5.1% out of all Submissions )
- [NeurIPS 2024] Chenhui Xu, Fuxun Yu, Maoliang Li, Zihao Zheng, Zirui Xu, Jinjun Xiong, and Xiang Chen. Infinite-Dimensional Feature Interaction, in Annual Conference on Neural Information Processing Systems, 2024.
- [ICML 2024] Chenhui Xu, Fuxun Yu, Zirui Xu, Nathan Inkawhich, and Xiang Chen. Out-of-Distribution Detection via Deep Multi-Comprehension Ensemble, in International Conference on Machine Learning, 2024.
- [ASP-DAC 2024] Chenhui Xu, Fuxun Yu, Zirui Xu, Chenchen Liu, Jinjun Xiong, and Xiang Chen. QuadraNet: Improving High-Order Neural Interaction Efficiency with Hardware-Aware Quadratic Neural Networks, Asia and South Pacific design automation conference, IEEE 2024. (Best Paper Nomination. 10 out of 482 Submissions)
- [TODAES] Ruiyang Qin, Dancheng Liu, Chenhui Xu, Zheyu Yan, Zhaoxuan Tan, Zixuan Pan, Zhenge Jia, Amir Nassereldine, Jiajie Li, Meng Jiang, Ahmed Abbasi, Jinjun Xiong, Yiyu Shi. Empirical Guidelines for Deploying LLMs onto Resource-constrained Edge Devices, ACM Transactions on Design Automation of Electronic Systems, 2025.
Preprints
- [arxiv 2024] Amir Nassereldine, Dancheng Liu, Chenhui Xu, Jinjun Xiong. PI-Whisper: An Adaptive and Incremental ASR Framework for Diverse and Evolving Speaker Characteristics, arXiv preprint arXiv:2406.15668, 2024.
- [arxiv 2024] Dancheng Liu, Amir Nassereldine, Ziming Yang, Chenhui Xu, Yuting Hu, Jiajie Li, Utkarsh Kumar, Changjae Lee, Jinjun Xiong. Large Language Models have Intrinsic Self-Correction Ability, arXiv preprint arXiv:2406.15673, 2024.
Education Background
- Pursuing Ph.D. in Computer Science Aug. 2026∼present
- M.S. in Research. in Computer Science and Engineering, Transfer to UMD Aug. 2024∼Aug. 2026
- Transfer to UB, Computer Engineering Aug. 2022∼May. 2024
- Bachelor of Sciences in Statistics Aug. 2017∼ Jun. 2021
Services
- Reviewer, Annual Conference on Neural Information Processing Systems (NeurIPS), 2025, 2026
- Reviewer, International Conference on Machine Learning (ICML), 2025, 2026
- Reviewer, International Conference on Artificial Intelligence and Statistics (AISTATS), 2025
- Reviewer, International Conference on Learning Representation (ICLR), 2025, 2026
- Reviewer, NeurIPS Workshop on MATH-AI, 2024
- Session Chair, IEEE Cloud Summit, 2022,2023
- Publication Committee, International Green and Sustainable Computing Conference, 2022, 2023
Contact Information
Davis Hall, Room 341
Computer Science & Engineering Department
University at Buffalo
Buffalo, NY, USA 14260
Email: cxu26@buffalo.edu
Phone: +1(703)973–0703
Global Visitors
Live map of visits from around the world.