Hi, I am Hongduan Tian, a fourth-year Ph.D. student at Trustworthy Machine Learning and Reasoning (TMLR) Group in Department of Computer Science, Hong Kong Baptist University, advised by Dr. Bo Han, and work closely with Dr. Feng Liu. Before that, I got my master degree from Nanjing University of Information Science and Technology (NUIST) and was fortunately supervised by Prof. Xiao-Tong Yuan and Prof. Qingshan Liu.

My research interests involves two aspects: (1) exploring the underlying mechanism of foundation model abilities (e.g., reasoning) and (2) improving the efficiency of foundation models. The goal of my research is to develop efficient foundation model agents based on trustworthy capabilities. Previously, I also explore test-time adaptation across various domains. Here are some representative works:

Trustworthy Foundation Model Agents: Exploring underlying mechanisms of agentic framework and building FM agentic frameworks.
[MAD-M$^2$] [LangPINN]
Test-time Adaptation with Limited Samples: Achieving good generalization performance on tasks from previous data or domains.
[MOKD] [CoPA] [IFR]

We are also organizing the TMLR Young Scientist Seminars and actively seeking researchers interested in sharing their work. If you would like to give a talk, we encourage you to reach out to us.

Please feel free to email me for research, collaborations, or a casual chat.

Email: cshdtian [at] comp.hkbu.edu.hk / hongduan.tian [at] gmail.com

📣 News

$\frak{2026.03}$: I will work as visiting student at RIKEN Imperfect Information Learning Team.
$\frak{2026.03}$: Our paper “Lang-PINN: From Language to Physics-Informed Neural Networks via a Multi-Agent Framework” is accepted by ICLR 2026 Workshop on AI with Recursive Self-Improvement and is highlighted as Spotlight.
$\frak{2026.01}$: Our paper “Multi-Agent Debate with Memory Masking” is accepted by ICLR 2026.
$\frak{2025.11}$: Our paper “Cross-domain Few-shot Classification via Invariant-content Feature Reconstruction” is accepted by IJCV.
$\frak{2024.09}$: Our paper “Mind the gap between prototypes and images in cross-domain finetuning” is accepted by NeurIPS 2024.
$\frak{2024.05}$: Our paper “MOKD: Cross-domain finetuning for few-shot classification via maximizing optimized kernel dependence” is accepted by ICML 2024.

📖 Educations

2023.09 - present, Ph.D. Student, TMLR Group, Hong Kong Baptist University (HKBU), advised by Dr. Bo Han.
2026.03 - 2026.06, Visiting Student, RIKEN AIP, advised by Dr. Takashi Ishida and Prof. Masashi Sugiyama.
2018.09 – 2021.06, M.Eng. in Automation, Nanjing University of Information Science and Technology (NUIST), advised by Prof. Xiao-Tong Yuan and Prof. Qingshan Liu.
2014.09 - 2018.06, B.E. in Automation, Nanjing University of Information Science and Technology (NUIST).

📝 Publications

✉️ Corresponding author.

Lang-PINN: From Language to Physics-Informed Neural Networks via a Multi-Agent Framework. [Spotlight]

Liangliang You, Xin He, Hongduan Tian, Bo Han, Ivor Tsang, Yew-Soon Ong.

[paper] [code] [slides] [poster] [CN-video] [EN-video]

Multi-Agent Debate with Memory Masking.

Hongduan Tian, Xiao Feng, Ziyuan Zhao, Xiangyu Zhu, Rolan Yan, Bo Han✉️.

[paper] [code] [slides] [poster] [CN-video] [EN-video]

Cross-domain Few-shot Classification via Invariant-content Feature Reconstruction.

Hongduan Tian, Feng Liu, Ka Chun Cheung, Zhen Fang, Simon See, Tongliang Liu, Bo Han✉️.

[paper] [code]

Mind the Gap Between Prototypes and Images in Cross-domain Finetuning.

Hongduan Tian, Feng Liu, Zhanke Zhou, Tongliang Liu, Chengqi Zhang, Bo Han✉️.

[paper] [code] [slides] [poster] [CN-video] [EN-video]

MOKD: Cross-domain Finetuning for Few-shot Classification via Maximizing Optimized Kernel Dependence.

Hongduan Tian, Feng Liu, Tongliang Liu, Bo Du, Yiu-ming Cheung, Bo Han✉️.

[paper] [code] [slides] [poster] [CN-video] [EN-video]

Meta-learning with network pruning.
Hongduan Tian✉️, Bo Liu, Xiao-Tong Yuan✉️, Qingshan Liu.

[paper] [code]

🎖 Awards

2026.05, Research Performance Award, HKBU CS Department.
2026.05, Silver Reviewer Award, ICML’26.
2024.11, Research Performance Award, HKBU CS Department.
2024.10, NeurIPS Scholar Award.

🏦 Project Grants

Research and Practice Innovation Program for Graduate Students in Jiangsu Province, SJCX20_0302, “Research on Theory and Practice of Optimization-based Meta-Learning”, 2020.04 - 2021.04

🎤 Invited Talks

2024.11, Mind the Gap Between Prototypes and Images in Cross-domain Finetuning @AI TIME, Online. [Video]
2024.06, MOKD: Cross-domain finetuning for few-shot classification via maximizing optimized kernel dependence @AI TIME, Online. [Video]

💻 Services

Conference Reviewer for ICML, NeurIPS, ICLR, AISTATS.
Journal Reviewer for TPAMI, TIP, TNNLS, TMLR, NEUNET.

🏫 Teaching

COMP7930 (G) Big Data Analytics, Sem. 2, 2025 - 2026
COMP7015 (G) Artificial Intelligence, Sem. 1, 2025 - 2026
COMP7250 (G) Machine Learning, Sem. 2, 2024 - 2025
COMP7180 (G) Quantitative Methods for Data Analytics and Artificial Intelligence, Sem. 1, 2024 - 2025
COMP7940 (G) Cloud Computing, Sem. 2, 2023 - 2024

🏢 Experiences

2024.11 - 2026.04, Remote Research Intern @WeChat, supervised by Xiangyu Zhu and Rolan Yan
2022.07 - 2025.12, Research Intern @NVIDIA NVAITC, host by Charles Cheung.
2024.06 - 2024.08, Research Intern @WeChat, host by Xiangyu Zhu and Rolan Yan
2023.07 - 2023.08, Research Intern @Alibaba.
2022.07 - 2023.05, Research intern @HKBU-TMLR Group, advised by Dr. Bo Han and Dr. Feng Liu.
2021.07 - 2022.07, Algorithm Engineer @ZTE Nanjing Research and Development Center.