Tzu-Lam Lee

Tzu-Lam Lee(Zilin Li)

李子林

BE Student
School of Information and Intelligent Science, Donghua University, Shanghai
tzulamlee[at]gmail[dot]com

About

四年前,我放棄自己最熱忱的化學,來到 Shanghai, 投身電腦科學領域。

四年後,我即將來到 Hung Hom——這個距離維多利亞港灣唔足一英里的地方——加入理大 DSAI,同時獲得香港 特別行政區大學教育資助委員會(UGC)資助,在 Prof. Irwin King 學生 Prof. Zixing Song 的指導下,繼續開展電腦輔助藥物設計(CADD)的工作,並完成後續哲學碩士的攻讀。

這對於我本人而言意義非凡。兜兜轉轉,彷彿一切又回到了原點。

三年間,我一直專注於人工圖神經網絡以及表徵學習方向的工作, 研究興趣涵蓋冗餘表徵纏結解離、因子降維和不確定性量化分析等。我亦接受過心理學與計算神經科學的系統學術訓練,學術脈絡上承繼威廉·馬克西米利安·馮特(Wilhelm Maximilian Wundt) 實驗心理學的量化研究傳統,堅持以實證與計算建模的路徑探索表徵學習的底層機制。

Research Interests

Scientific Machine Learning: graph neural networks and generative models for scientific simulation, prediction, and design.
Uncertainty Quantification: uncertainty estimation, calibration, and decision-making under uncertainty.
Multi-Agent Learning: cooperative decision-making, emergent coordination, and distributed learning.
Multimodal and Efficient Learning: multimodal representations, vision-language models, and resource-efficient deep learning.

Education

2023–2027 (Expected)
Donghua University
Bachelor of Engineering · Shanghai
2024–2025
Shanghai Jiao Tong University
Intern · Shanghai

Publications

NeurIPS 2026
NeurIPS Decentralized Coupled Representation Learning
Zilin Li (first author) · Accepted · Poster
ICML 2026 · EIML Workshop
ICML PIERCE: Triadic Bayesian Reasoning for Establishing Epistemic Boundaries
Zilin Li (first author) · Accepted

Advisor

Ge Zhang — Lecturer, Donghua University

Fellowship