I am a first-year student in College of Engineering at Boston University (BU), advised by Prof. Yigong Hu.

I earned my master’s degree (MPhil.) in Data Science and Analytics Thrust (DSA) at The Hong Kong University of Science and Technology (Guangzhou), supervised by Prof. Zeyi Wen and co-supervised by Prof. Xinyu Chen. Before that, I received my bachelor’s degree (B.Eng.) at Guangzhou University, where I was fortunate to work closely with Prof. Xue-Feng Yuan during my junior and senior years.

I used to study data sparsity problems in sequential recommender system and applications of low-rank techniques in large language models, published 6 papers (my google scholar: ) on these subjects during my master’s research. Now, I am more interested in system reliability for my Ph.D. research, focusing on building a reliable and fast system.

🔥 News

  • 2025.07:  🎓 Received my master’s degree (MPhil. in Data Science and Analytics) at The Hong Kong University of Science and Techonlogy.

📝 Publications

System

ArXiv
sym

gigiProfiler: Diagnosing Performance Issues by Uncovering Application Resource Bottlenecks

Yigong Hu, Haodong Zheng, Yicheng Liu, Dedong Xie, Youliang Huang, Baris Kasikci

  • Introduce OmniResource Profiling to track both system-level and application-level resources to comprehensively dignose resource bottlenecks.
  • Using hybrid LLM-static analysis approach to identify application resources with LLMs and analyze performance bottlenecks with buggy execution.

Large Language Models

📊 Recommender System

WWW 2025 (Oral)
sym

When Large Vision Language Models Meet Multimodal Sequential Recommendation: An Empirical Study

Peilin Zhou, Chao Liu, Jing Ren, Xinfeng Zhou, Yueqi Xie, Meng Cao, Zhongtao Rao, You-Liang Huang, Dading Chong, Junling Liu, Jae Boum Kim, Shoujin Wang, Raymond Chi-Wing Wong, Sunghun Kim

  • Introduces MSRBench, the first benchmark to systematically evaluate the integration of Large Vision Language Models (LVLMs) in multimodal sequential recommendation.
  • Compares five integration strategies (recommender, item enhancer, reranker, and two combinations), identifying reranker as the most effective.
  • Constructs the enhanced dataset Amazon Review Plus, with LVLM-generated image descriptions to support more flexible item modeling.

📚 Others

Computers & Graphics
sym

StyleTerrain: A novel disentangled generative model for controllable high-quality procedural terrain generation, Computers & Graphics, Volume 116, Pages 373-382

You-Liang Huang, Xue-Feng Yuan

  • It introduces disentangled representation learning into Generative Adversarial Network (GAN)-based terrain modeling methods.
  • Disentangling latent space to achieve controllable DEM generation with diverse landscape features.

💁 Professional Services

  • AAAI’26, AAAI’26 (AIA), Program Committee
  • ICLR’25, Reviewer
  • Eurographics’24, Reviewer

🎖 Honors and Awards

  • 2023.05 Dean’s List and First Price Studentship, Guangzhou University.
  • 2022.11 Dean’s List and First Price Studentship, Guangzhou University.
  • 2021.10 Dean’s List and First Price Studentship, Guangzhou University.
  • 2021.08 The third price in Computer System Development Capability Competition.
  • 2020.11 Dean’s List and First Price Studentship, Guangzhou University.

📖 Educations

  • 2025.09 - Present, Ph.D. student in Computer Engineering, Boston University.
  • 2023.09 - 2025.07, MPhil. in Data Science and Analytics, CGA: 3.66(4.3), Hong Kong University of Science and Technology.
  • 2019.09 - 2023.06, B.Eng. in Network Engineering, GPA: 3.87(4.0), Rank: 2nd(163), Guangzhou University.

💻 Internships

  • 2023.01 - 2023.10, Research Assistant, DeepSE Lab, The Hong Kong University of Science and Technology, China.
  • 2021.11 - 2023.06, Research Assistant, Institute of System Rheology, Guangzhou University, China.

🗺️ Visitor Map