Jingxuan(Jasper) Wu

Learning to Learn

Hi! I’m Jingxuan Wu (吴竟瑄). I am a master’s student in Statistics & Operations Research at the University of North Carolina at Chapel Hill, and I received my B.S. from The Chinese University of Hong Kong, Shenzhen. Since my undergraduate years at CUHK-Shenzhen, I have been fortunate to be advised by Prof. Guanting Chen at UNC–Chapel Hill, with whom I continue to collaborate during my master's studies. In 2025, I began working with Prof. Xingrui Yu and Prof. Ivor Tsang at A*STAR. Previously, I also worked with Prof. Jianfeng Mao.

I am currently seeking 2026 summer research opportunities and Ph.D. positions for Fall 2027. I’m also open to research collaborations. If you’re interested in my work, please contact me directly via the email listed below.

Email:

Learning to Learn
News

09/2026: Two papers CINC and TAPS were accepted to NeurIPS 2026. See you in Sydney 🇦🇺 and Atlanta 🇺🇸!

05/2026: Our paper TAPS was accepted to the ICML 2026 Workshop on SPIGM.

04/2026: Two papers OSCAR and FM-IRL were accepted to ICML 2026. See you in Seoul! 🇰🇷

Research

My research aims to build controllable generative systems through the integration of diffusion/flow-based models and multi-agent systems: (1) Generative Models—enabling diverse, fine-grained controllable generation under limited supervision through both training-free inference-time control and RL-based training; (2) Agentic Systems—designing multi-agent systems that interact, coordinate, and adapt through feedback.

Selected Publications

Scroll to view more. (* indicates equal contribution)

Preprint MemFold framework overview
MemFold: Learning Compact Soft Memory for Long-Context Personalization via On-Policy Optimization

Jingxuan Wu*, Yuzhe Yang*, Yiqiao Huang, Chengzhi Liu, Qingni Wang, Chengxuan Qian, Shutong Wu, Jiawei Zhang, Xin Eric Wang

Learns compact soft memories for long-context personalization through on-policy optimization.

Paper / Code / Project
ICML 2026 OSCAR overview
Letting Trajectories Spread: Quality-Preserving Control for Diverse Flow Matching

Jingxuan Wu*, Zhenglin Wan*, Xingrui Yu, Yuzhe Yang, Bo An, Ivor Tsang

Improves sample diversity in flow matching while preserving generation quality.

Paper / Code
ICML 2026 AD-OPD framework overview
Adversarial Dual On-Policy Distillation from Expressive Teacher

Zhenglin Wan*, Jingxuan Wu*, Xingrui Yu, Chubin Zhang, Mingcong Lei, Bo An, Ivor Tsang

Distills expressive teachers into efficient policies through adversarial dual on-policy learning.

Paper / Code
NeurIPS 2026 TAPS overview
Denoising Time Matters: Diverse Generation in Diffusion Language Models

Jingxuan Wu, Zhenglin Wan, Xingrui Yu, Yuzhe Yang, Yiqiao Huang, Ivor Tsang, Yang You

Improves diversity in diffusion language models through time-annealed perturbations while preserving generation quality.

Paper / Code / Project

See all publications →
Experiences

University of Wisconsin–Madison 2026.5 – Present
Research Intern
Advised by Prof. Jiawei Zhang.
University of North Carolina at Chapel Hill 2024.1 – Present
Research Assistant & M.S. Student (Statistics & Operations Research)
Advised by Prof. Guanting Chen.
Agency for Science, Technology and Research (A*STAR) 2025.6 – Present
Research Intern
Advised by Prof. Xingrui Yu and Prof. Ivor Tsang.
School of Data Science, CUHK–Shenzhen 2023.11 – 2025.5
Undergraduate Research Assistant
Advised by Prof. Jianfeng Mao.
Education
University of North Carolina at Chapel Hill 2025 – Present
M.S. in Statistics and Operations Research
The Chinese University of Hong Kong, Shenzhen 2021 – 2025
B.S. in Data Science and Big Data Technology
First Class Honors