I build learning systems that connect long-horizon objectives with reliable credit assignment and memory.
About
Research Perspective
Connecting learning dynamics, memory, and intelligent behavior.
My work focuses how artificial neural systems can preserve useful experience, assign credit across long horizons, and form generalisable representations. I approach these questions through recommender systems, associative memory, representation learning, and computational neuroscience.
Before, I earned my PhD in Artificial Intelligence at the State Key Lab of Brain-Machine Intelligence, ZJU. My dissertation investigated how sparse latent codes can support recognition, memory formation, and memory-conditioned computation.
01
Recognition & Service
Selected honors and academic contributions.
Honors
- Inaugural Tencent Project UP Talent Initiative 入选首届"青云计划"
- China Mobile Project Golden Seed Talent Initiative
- Award of Honor for Graduate of Zhejiang University
- Outstanding Graduate Student of CCNT, Zhejiang University
- Graduate of Merit / Triple A Graduate, Zhejiang University
Academic Service
Served for top-tier tracks&journals including NeurIPS, ICLR, ICML, IJCAI, ICASSP, Proceedings of the IEEE, Nature Communications, IEEE TPAMI, IEEE TNNLS, Neural Networks, and related venues.
International Exchange
- Chinese University of Hong Kong, 2016
- Technical University of Munich, 2014
- New York University, 2012
02
Selected Research
Work on memory, credit assignment, and learning systems.
Research on a Spiking Generative Model of Spatiotemporal Memory Construction and Computation
A unified model which supports multimodal input, associative memory, and learned by temporal credit assignment in sparse neural models.
Temporal Conditioning Spiking Latent Variable Models of the Neural Response to Natural Visual Scenes
A memory-conditioned latent variable model for predicting ultra-long neural responses to natural visual scenes.
Exploiting Noise as a Resource for Computation and Learning in Spiking Neural Networks
A theoretical and empirical framework that leverages neuronal noise as learning signals.
Dual Memory Model for Experience-once Task-incremental Lifelong Learning
A complementary-memory architecture designed for rapid, experience-once continual learning.
Successive POI Recommendation via Brain-inspired Spatiotemporal Aware Representation
A model that jointly captures spatial and temporal context for successive point-of-interest recommendation.
Bioimaging of Dissolvable Microneedle Arrays: Challenges and Opportunities
A review of imaging methods, measurement challenges, and opportunities for medical bioimages.
03
Experience
Applied research across recommender systems, machine intelligence, and computational neuroscience.
- 2024.08 - Present
- 2022.07 - 2024.07
Liangzhu Laboratory, Medical Center - The First Affiliated Hospital, ZJU School of Medicine
Research Intern
Representation learning and associative memory.
- 2020.09 - 2024.06
The State Key Lab of Brain-Machine Intelligence, ZJU
PhD Candidate, Zhejiang University
- 2020.07 - 2021.06