Haoyu Huang

I am a Ph.D. student of HKUST at KnowComp, advised by Prof. Yangqiu Song.

Before that, I got my master degree from CUHK, advised by Prof. James Cheng. I also got my B.E. degree from BJTU, advised by Prof. Huaiyu Wan. I also worked as a Research Assistant in the Center of Machine Learning Research of PKU, advised by Prof. Wentao Zhang.

My research interests include some key aspects of agentic knowledge systems:

• In-Parameter Learning (IPL) for LLMs: AtlasKV.
• Data Agents: DeepRefine, HiRAG, NGDBench.
• Knowledge Graphs: STKGRec, STKG-PLM, GraphJudge, AutoSchemaKG, AutoGraph-R1.

Open-Source Projects

• DeepRefine is a skill that plugs into agent workflows and use a single command /deeprefine in your agent CLI to refine and evolve your LLM-Wiki at test time.
  

• HiRAG is an open-source project that introduces unsupervised hierarchical KG indexing and a novel bridging mechanism for effective knowledge integration, significantly advancing the state-of-the-art in RAG systems.
  

• AutoSchemaKG is a novel framework for automatic knowledge graph construction that combines schema generation via conceptualization.
  

Selected Preprints and Publications

• DeepRefine: Agent-Compiled Knowledge Refinement via Reinforcement Learning

  Haoyu Huang, Jiaxin Bai, Shujie Liu, Yang Wei, Hong Ting Tsang, Yisen Gao, Zhongwei Xie, Yufei Li, Yangqiu Song#

  In Submission
  [paper] [code]

• AtlasKV: Augmenting LLMs with Billion-Scale Knowledge Graphs in 20GB VRAM

  Haoyu Huang, Hong Ting Tsang, Jiaxin Bai, Xi Peng, Gong Zhang, Yangqiu Song#

  In International Conference on Learning Representations (ICLR), 2026
  [paper] [code] [Project Page]

• AutoGraph-R1: End-to-End Reinforcement Learning for Knowledge Graph Construction

  Hong Ting Tsang, Jiaxin Bai, Haoyu Huang, Qiao Xiao, Tianshi Zheng, Baixuan Xu, Shujie Liu, Yangqiu Song#

  In Annual Meeting of the Association for Computational Linguistics (ACL), 2026, Main Conference
  [paper] [code]

• Retrieval-Augmented Generation with Hierarchical Knowledge

  Haoyu Huang*, Yongfeng Huang*, Junjie Yang, Zhenyu Pan, Yongqiang Chen, Kaili Ma, Hongzhi Chen, James Cheng#

  In Empirical Methods in Natural Language Processing (EMNLP), 2025, Findings
  [paper] [code]

• Can LLMs be Good Graph Judge for Knowledge Graph Construction?

  Haoyu Huang, Chong Chen, Zeang Sheng, Yang Li, Wentao Zhang#

  In Empirical Methods in Natural Language Processing (EMNLP), 2025, Main Conference
  [paper] [code]

• Next-POI Recommendation via Spatial-Temporal Knowledge Graph Contrastive Learning and Trajectory Prompt

  Wei Chen*, Haoyu Huang*, Tianyi Wang, Zhiyu Zhang, Shengnan Guo, Youfang Lin, Huaiyu Wan#

  In IEEE Transactions on Knowledge and Data Engineering (TKDE), 2025
  [paper] [code]

Education

• Ph.D. in CSE at HKUST, Hong Kong SAR, China 2025-2029(expected)

• M.Sc. in CSE at CUHK, Hong Kong SAR, China 2024-2025

• B.E. in CS at BJTU, Beijing, China 2019-2023

Selected Awards

• 🏆 Huawei Spark Award 2024

  🏆 Second Prize of Huawei Cup CPMCM (National) 2024

Academic Service

• Conference Reviewer: ARR Feb 2025, ARR May 2025, ICML 2026, NeurIPS 2026, ARR May 2026 (EMNLP).

  Journal Reviewer: IEEE TKDE.

  Conference Volunteer: EMNLP 2025.

Teaching Assistant

• 2026 Spring:

  Natural Language Processing. CSIT5520

  Big Data Mining and Management/Advanced Data Miningfor Risk Management and Business Intelligence. COMP4332/RMBI4310