PhD student at HKUST-KnowComp

Neural graph database, reasoning, and geometric deep learning

Yisen Gao

I work on neural graph databases, abductive and deductive reasoning in knowledge graphs, AI for mathematics, geometric deep learning, and recursive self-improvement for agents.

Email Google Scholar DBLP

About

I am a PhD student in HKUST-KnowComp supervised by Prof. Yangqiu Song. I received my bachelor's degree in Artificial Intelligence from Beihang University in June 2025. I am currently an intern in Tencent's Qingyun Program.

Before joining HKUST, I interned at Magic Group for two years, supervised by Prof. Jianxin Li, Prof. Qingyun Sun, and Prof. Xingcheng Fu. I received support from the Beijing Natural Science Foundation in 2024 and have served as a reviewer or program committee member for IJCAI, ICML, ARR, NeurIPS, AAAI, and ICLR.

I am open to collaborations around graph learning, reasoning, and neural graph data management.

Research Interests

Agent Recursive Self-Improvement

Building agents that learn from experience and evolve their knowledge and capabilities over time.

Neural Graph Databases

Building foundation models and systems for understanding graph databases.

Abductive Reasoning

Studying the fusion of deductive and abductive reasoning over knowledge graphs.

AI for Mathematics

Exploring autonomous AI mathematicians and reasoning-oriented learning.

Geometric Deep Learning

Understanding data through Riemannian geometry and hyperbolic representation learning.

Open Source Projects

Knosis

An open-source knowledge layer for agents that extracts and maintains evidence-backed knowledge for reuse across tasks. GitHub

Selected Publications

All Publications

preprint

Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds.

Yisen Gao, Yue Guo, Qing Zong, Yiwen Guo, Yangqiu Song.

paper code

ICLR 2026

Controllable Logical Hypothesis Generation for Abductive Reasoning in Knowledge Graphs.

Yisen Gao, Jiaxin Bai, Tianshi Zheng, Ziwei Zhang, Qingyun Sun, Xingcheng Fu, Jianxin Li, Yangqiu Song.

paper

preprint

KGPFN: Unlocking the Potential of Knowledge Graph Foundation Model via In-Context Learning.

Yisen Gao, Jiaxin Bai, Haoyu Huang, Zhongwei Xie, Yufei Li, Hong Ting Tsang, Sirui Han, Yangqiu Song.

paper

preprint

HypoAgent: An Agentic Framework for Interactive Abductive Hypothesis Generation over Knowledge Graphs.

Yisen Gao, Yixi Cai, Tianshi Zheng, Jiaxin Bai, Yangqiu Song.

paper

WWW 2026 Oral

Unifying Deductive and Abductive Reasoning in Knowledge Graphs with Masked Diffusion Model.

Yisen Gao, Jiaxin Bai, Yi Huang, Xingcheng Fu, Qingyun Sun, Yangqiu Song.

paper

WWW 2026

Graph Diffusion Evolution Model for Multi-Conditional Molecular Generation.

Xingcheng Fu, Lingyun Liu, Yisen Gao, Tianyu Chen, Qingyun Sun, Jianxin Li, Xianxian Li.

paper

WWW 2026

Towards Geometry-Consistent Federated Graph Learning.

Yuecen Wei, Zhiyu Zhuang, Yisen Gao, Xingcheng Fu, Qingyun Sun, Ziwei Zhang, Tianyu Wo, Chunming Hu.

paper

AAAI 2026

Towards LLM-Empowered Knowledge Tracing via LLM-Student Hierarchical Behavior Alignment in Hyperbolic Space.

Xingcheng Fu, Shengpeng Wang, Yisen Gao, Xianxian Li, Chunpei Li, Qingyun Sun, Dongran Yu.

paper

AAAI 2026

Is the Information Bottleneck Robust Enough? Towards Label-Noise Resistant Information Bottleneck Learning.

Yi Huang, Qingyun Sun, Yisen Gao, Haonan Yuan, Xingcheng Fu, Jianxin Li.

paper

NeurIPS 2026

NGDB-Zoo: Towards Efficient and Scalable Neural Graph Databases Training.

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

paper

EMNLP 2026

NGDBench: Towards Neural Graph Data Management.

Yufei Li^*, Yisen Gao^*, Jiaxuan Xiong, Jiaxin Bai, Shijie Zhong, Haoyu Huang, Zhongwei Xie, Hong Ting Tsang, Yangqiu Song.

paper

DEBull 2025

Top Ten Challenges Towards Agentic Neural Graph Databases.

Jiaxin Bai, Zihao Wang, Yukun Zhou, Hang Yin, Weizhi Fei, Qi Hu, Zheye Deng, Jiayang Cheng, Tianshi Zheng, Hong Ting Tsang, Yisen Gao, Zhongwei Xie, Yufei Li, Lixin Fan, Binhang Yuan, Wei Wang, Lei Chen, Xiaofang Zhou, Yangqiu Song.

paper

AAAI 2025

GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts.

Zihao Guo, Qingyun Sun, Haonan Yuan, Xingcheng Fu, Min Zhou, Yisen Gao, Jianxin Li.

paper

AAAI 2025

Discrete Curvature Graph Information Bottleneck.

Xingcheng Fu, Jian Wang, Yisen Gao, Qingyun Sun, Haonan Yuan, Jianxin Li, Xianxian Li.

paper

AAAI 2025 Oral

Bi-Directional Multi-Scale Graph Dataset Condensation via Information Bottleneck.

Xingcheng Fu^*, Yisen Gao^*, Beining Yang, Yuxuan Wu, Haodong Qian, Qingyun Sun, Xianxian Li.

paper

CVPR 2025

Galaxy Walker: Geometry-aware VLMs For Galaxy-scale Understanding.

Tianyu Chen, Xingcheng Fu, Yisen Gao, Haodong Qian, Yuecen Wei, Kun Yan, Haoyi Zhou, Jianxin Li.

paper

NeurIPS 2025

Toward a Unified Geometry Understanding: Riemannian Diffusion Framework for Graph Generation and Prediction.

Yisen Gao, Xingcheng Fu, Qingyun Sun, Jianxin Li, Xianxian Li.

paper

DAI 2025

SIGMA: A Dual-Agent Reinforcement Learning-Optimized Framework for Graph Classification.

Xinbang Cheng, Haodong Qian, Yisen Gao, Xingcheng Fu, Rongye Shi.

paper

ICML 2024

Hyperbolic Geometric Latent Diffusion Model for Graph Generation.

Xingcheng Fu, Yisen Gao, Yuecen Wei, Qingyun Sun, Hao Peng, Jianxin Li, Xianxian Li.

paper

News

2026

Three papers accepted by NeurIPS 2026.

NGDB-Zoo, time-series generation from a frequency-domain perspective, and LLM-assisted decision tree generation in hyperbolic space. Details of the latter two works are forthcoming.

2026

One paper accepted by EMNLP 2026.

NGDBench: Towards Neural Graph Data Management.

2026

One paper accepted by ICLR 2026.

Controllable logical hypothesis generation for abductive reasoning in knowledge graphs.

2026

Three papers accepted by WWW 2026.

Topics include masked diffusion for reasoning, molecular generation, and geometry-consistent federated graph learning.

2026

Two papers accepted by AAAI 2026.

Topics include hyperbolic knowledge tracing and robust information bottleneck learning.

2025

One paper accepted by NeurIPS 2025.

Riemannian diffusion framework for graph generation and prediction.

2025

One paper accepted by CVPR 2025.

Geometry-aware VLMs for galaxy-scale understanding.

2025

Three papers accepted by AAAI 2025.

Topics include Riemannian experts, graph information bottleneck, and graph dataset condensation.

2024

One paper accepted by ICML 2024.

Hyperbolic geometric latent diffusion model for graph generation.

Beyond Research

Cooking: I enjoy delicious food and exploring diverse, novel taste combinations.
Music: I listen to many genres and am a fan of Yichun Shan.

Collaboration: Please contact me if you are interested in graph learning, reasoning, or neural graph databases.
Academic service: Reviewer or PC member for IJCAI, ICML, ARR, NeurIPS, AAAI, and ICLR.