Algorithm Engineer @ Kuaishou ยท RecSys & Agent
"All models are wrong, but some are useful."
I'm Mingjie Sun (ๅญๆๆฐ), an Algorithm Engineer at Kuaishou Technology, focusing on advertising recommendation models. My current research interests include Generative Recommendation and Agent for RecSys. Previously, I conducted research in cybersecurity, advised by Prof. Yong Liao and Prof. Haiyong Xie at USTC.
Working on advertising recommendation models. Current research focuses on Generative Recommendation and Agent for RecSys, exploring the integration of large language models with recommendation systems.
Conducted research in network security and published multiple academic papers (see Google Scholar).
Built a solid foundation in computer science and communications engineering.
Align3GR: Unified Multi-Level Alignment for LLM-based Generative Recommendation
AAAI 2026 OralGenerative Recommendation for Large-Scale Advertising
KDD 2026DAS: Dual-Aligned Semantic IDs Empowered Industrial Recommender System
CIKM 2025 OralSweetTok: Semantic-Aware Spatial-Temporal Tokenizer for Compact Video Discretization
ICCV 2025Spatial-Temporal Attention Network for Crime Prediction with Adaptive Graph Learning
ICANN 2022Identification of Critical Nodes in Dynamic Systems Based on Graph Convolutional Networks
ICUS 2020* Equal contribution (co-first author)
I'm always open to interesting conversations and new opportunities.