Haoyu Huang
Ph.D. Student
HKUST
Email: hhuangcp [at] connect (dot) ust (dot) hk
Ph.D. Student
HKUST
Email: hhuangcp [at] connect (dot) ust (dot) hk
I am a Ph.D. student of HKUST, advised by Prof. Yangqiu Song. I am also a research intern at Microsoft Research Aisa.
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 intern in the DCAI group of PKU, advised by Prof. Wentao Zhang. My past research focuses on RAG, Knowledge Graphs, Recommendations.
My current research focuses on Agent Post-Training, LLM Memory.
One paper accepted to NeurIPS 2026. 2026.9
One paper accepted to EMNLP 2026 Findings. 2026.8
Two papers accepted to ACL 2026. 2026.4
One paper accepted to ICLR 2026. 2026.1
Two papers accepted to EMNLP 2025. 2025.8
One paper accepted to IEEE TKDE. 2025.2
DeepRefine. 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. An easy-to-use GraphRAG system that introduces hierarchical indexing and a novel bridging mechanism for effective knowledge integration.
Towards In-Parameter Memory Augmentation for Large Language Models
In Submission
DeepRefine: Agentic Knowledge Refinement via Reinforcement Learning
In Submission
AtlasKV: Augmenting LLMs with Billion-Scale Knowledge Graphs in 20GB VRAM
In International Conference on Learning Representations (ICLR), 2026
AutoGraph-R1: End-to-End Reinforcement Learning for Knowledge Graph Construction
In Annual Meeting of the Association for Computational Linguistics (ACL), 2026, Main Conference
Retrieval-Augmented Generation with Hierarchical Knowledge
In Empirical Methods in Natural Language Processing (EMNLP), 2025, Findings
Can LLMs be Good Graph Judge for Knowledge Graph Construction?
In Empirical Methods in Natural Language Processing (EMNLP), 2025, Main Conference
Next-POI Recommendation via Spatial-Temporal Knowledge Graph Contrastive Learning and Trajectory Prompt
In IEEE Transactions on Knowledge and Data Engineering (TKDE), 2025MOSAIC: An Omni-Modal Benchmark for Long-Term Memory and Cross-Modal Reasoning in Conversational Agents
In Submission [paper]RadixFlow: Filling Execution Bubbles in Prefix-Cached Hybrid LLM Serving
In Submission [paper]SoFT: Soft Targets for Generalizable Fine-Tuning
In Submission [paper]ACE-GraphRAG: Agentic Context Engineering for Hierarchical GraphRAG
In Submission [paper]AutoSchemaKG: Autonomous knowledge graph construction through dynamic schema induction from web-scale corpora
In Annual Meeting of the Association for Computational Linguistics (ACL), 2026, Main Conference [paper] [code]PatchWorld: Gradient-Free Optimization of Executable World Models
In Submission [paper] [code]KGPFN: Unlocking the Potential of Knowledge Graph Foundation Model via In-Context Learning
In Submission [paper] [code]NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training
In Conference on Neural Information Processing Systems (NeurIPS), 2026 [paper]NGDBench: Towards Neural Graph Data Management
In Empirical Methods in Natural Language Processing (EMNLP), 2026, Findings [paper] [code]Building and exploiting spatial–temporal knowledge graph for next POI recommendation
In Knowledge-Based Systems (KBS), 2022 [paper] [code]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
🏆 Huawei Spark Award 2024
Conference Reviewer or PC member: ARR Feb 2025 (ACL 2025), May 2025 (EMNLP 2025), May 2026 (EMNLP 2026), Aug 2026, Oct 2026 (NAACL 2026), ICML 2026, NeurIPS 2026, CIKM 2026, ICLR 2027.
Journal Reviewer: IEEE TKDE.
Conference Volunteer: EMNLP 2025.
2026 Spring:
Natural Language Processing. CSIT5520
Big Data Mining and Management / Advanced Data Miningfor Risk Management and Business Intelligence. COMP4332/RMBI4310
2026 Fall:
Advanced Machine Learning with Graphs. COMP5222