Incoming MPhil student at HKUST (Guangzhou) with Prof. Hui Xiong (2027) · B.Eng. in Intelligent Science & Technology at Zhejiang University of Technology (expected June 2027) · research intern at AIR, Tsinghua University · AI research intern at Hithink RoyalFlush.
- AI for Science — neural operators for PDEs, and spatial–spectral fusion for hyperspectral and remote-sensing imaging.
- Autonomous Driving — world models for end-to-end driving: predicting how a scene unfolds, and scoring candidate trajectories with it.
- Multimodal Models — vision–language–action policies and omni-modal foundation models, including memory for long-horizon robot tasks.
| Venue | Paper | Links |
|---|---|---|
| Hyperbolic Neural Operator · first author | PDF · Project · Code | |
| S³O: Selective Spatial–Spectral Operator for Cross-Scale Fusion · co-first author | Paper · Code | |
| Asymmetric StarFus: Learning Incoherent Measurements for Semantics-Aware Spatial–Spectral Fusion · first author | Paper · Code | |
| GTR-Mamba: Geometry-to-Tangent Routing Mamba for Hyperbolic POI Recommendation · co-first author | arXiv | |
| Solving Spatial–Spectral Fusion with Latent Spectral Operators | Paper · Project · Code | |
| S²TA-Fuse: Semantic-Superpixel Tokenized Attention for Spatial–Spectral Fusion | Paper | |
| Mag-Mamba: Modeling Coupled Spatiotemporal Asymmetry for POI Recommendation | arXiv | |
| RoboMemArena: A Comprehensive and Challenging Robotic Memory Benchmark | arXiv · Project · Code |
- National First Prize — National Undergraduate Intelligent Car Race, 2025
- National Second Prize — National Undergraduate Electronic Design Contest, 2025
- National First Prize — China Robot Competition & RoboCup China, 2024
Figures and the full story on the homepage. The banner is built by assets/make_banner.py.

