Hao Yan颜浩

Sherirto · Ph.D. student

About me

Central South University
Graduating in June 2027

My name is Hao Yan (颜浩), and I also go by Sherirto. I am a Ph.D. student in Computer Science at Central South University, under the supervision of Prof. Senzhang Wang. I expect to graduate in June 2027.

From September 2025 to August 2026, I participated in the Joint Ph.D. Supervision Scheme at The Hong Kong Polytechnic University as a fully funded Research Assistant, supervised by Prof. Chengqi Zhang. I also receive research supervision from Prof. Shirui Pan.

I received my B.Sc. in Mathematics and Applied Mathematics from Hefei University of Technology in 2021. Previously, I was a research intern in the Social Computing Group at Microsoft Research Asia (July 2022–January 2023), mentored by Chaozhuo Li, and a remote research intern at Microsoft STCA (February 2023–January 2024), mentored by Weihao Han.

My research interests include:

Graph representation learning and graph–language models
Multimodal learning and robustness
Large language models, agents, and LLM reasoning
Looped Transformers and iterative computation

I am extending my research toward LLM reasoning, agent systems, and looped Transformer architectures.

Open to opportunities. I am seeking suitable Research Assistant and Algorithm Engineer positions, with interests in LLMs, agents, and reasoning. I expect to graduate in June 2027.

First-author research

5 first-author papers · 14 publications in total

Venue ratings follow the CCF 2026 catalogue.

A six-node multimodal graph and an expanded node showing a bird photograph with its text description. — Figure 1 · arXiv v2

KDD 2025CCF-A

When Graph Meets Multimodal: Benchmarking and Meditating on Multimodal Attributed Graph Learning

Hao Yan, Chaozhuo Li, Jun Yin, et al.

Five multimodal graph datasets and a unified evaluation of GNN-as-Predictor and VLM-as-Predictor, examining how domain, modality imbalance, and graph retrieval affect prediction.

Read paper View code

Benchmarking language and graph models for text-attributed graph learning.

NeurIPS 2023CCF-A

A Comprehensive Study on Text-attributed Graphs: Benchmarking and Rethinking

Hao Yan, Chaozhuo Li, Ruosong Long, et al.

A systematic benchmark for evaluating language models and graph neural networks on text-attributed graphs.

Read paper

Generic and task-specific language models meet graph representation learning.

WSDM 2025 · OralCCF-B

MoKGNN: Boosting Graph Neural Networks via Mixture of Generic and Task-Specific Language Models

Hao Yan, Chaozhuo Li, Jun Yin, et al.

Combining generic and task-specific language models to improve representation learning on text-attributed graphs.

View code

Balancing semantic consistency and augmentation diversity through cross-scale graph contrastive learning.

ACM TKDD 2025CCF-B

Have Our Cake and Eat It: Augmentation Diversity and Semantic Consistency Balanced Graph Contrastive Learning

Hao Yan, Senzhang Wang, Chaozhuo Li, Jun Yin, Philip S. Yu, Jianxin Wang

Combining semantic subgraph extraction and cross-scale mix-up to balance semantic consistency with augmentation diversity.

Read paper

Within-level and cross-level contrastive learning over a hierarchy of pooled graphs.

ECML-PKDD 2023CCF-B

Hierarchical Graph Contrastive Learning

Hao Yan, Senzhang Wang, Jun Yin, et al.

Learning graph representations with contrastive objectives both within and across levels of a graph hierarchy.

Read paper

More Publications 9 collaborative works

ICML 2026 · AcceptedCCF-A

Rel-MOSS: Towards Imbalanced Relational Deep Learning on Relational Databases

Jun Yin, Peng Huo, Bangguo Zhu, Hao Yan, et al.

Read paper

ICLR 2026CCF-A

Paradigm Shift of GNN Explainer from Label Space to Prototypical Representation Space

Jun Yin, Senzhang Wang, Ziluowen Luo, Peng Huo, Hao Yan, et al.

Read paper

CIKM 2026 · AcceptedCCF-B

PS²: Parameterized Control for Fine-Grained Student Proficiency Simulation

Ruochen Liu, Zhiyuan Wen, Hao Yan, Jun Yin, Senzhang Wang, Jiannong Cao

Read paper

IJCAI 2026CCF-B

One-Turn Knockout: Traceable and Editable Proxy Unlearning under Asymmetric Access Constraints

Ziluowen Luo, Jun Yin, Hao Yan, Ruochen Liu, Ming Cheng, Senzhang Wang

Read paper

WWW 2025 · OralCCF-A

Unleash LLMs Potential for Sequential Recommendation by Coordinating Dual Dynamic Index Mechanism

Jun Yin, Zhengxin Zeng, Mingzheng Li, Hao Yan, et al.

NeurIPS 2023CCF-A

Train Once and Explain Everywhere: Pre-training Interpretable Graph Neural Networks

Jun Yin, Chaozhuo Li, Hao Yan, Jianxun Lian, Senzhang Wang

Read paper

AAAI 2023CCF-A

WSiP: Wave Superposition Inspired Pooling for Dynamic Interactions-Aware Trajectory Prediction

Renzhi Wang, Senzhang Wang, Hao Yan, Xiang Wang

Read paper

ICLR 2023 · OralCCF-A

Learning on Large-scale Text-attributed Graphs via Variational Inference

Jianan Zhao, Meng Qu, Chaozhuo Li, Hao Yan, et al.

Read paper

SDM 2023CCF-B

Adversarial Hard Negative Generation for Complementary Graph Contrastive Learning

Senzhang Wang, Hao Yan, Jinlong Du, et al.

Read paper

Education

Sep. 2021–Jun. 2027 (expected)

Central South University

Combined M.Sc./Ph.D. in Computer Science and Technology

Advisor: Prof. Senzhang Wang.

Sep. 2025–Aug. 2026

The Hong Kong Polytechnic University

Joint Ph.D. Supervision Scheme · funded Research Assistant

Joint supervisor: Prof. Chengqi Zhang. Additional research supervision: Prof. Shirui Pan.

Sep. 2017–Jun. 2021

Hefei University of Technology

B.Sc. in Mathematics and Applied Mathematics

Research experience

Feb. 2023–Jan. 2024 · Remote

Microsoft STCA

Research Intern

Led systematic benchmarking of text-attributed graph learning, resulting in a first-author NeurIPS 2023 paper. Mentor: Weihao Han.

Jul. 2022–Jan. 2023

Microsoft Research Asia

Research Intern · Social Computing Group

Implemented core GLEM model code and conducted large-scale text-attributed graph experiments. Co-authored an ICLR 2023 oral paper. Mentor: Chaozhuo Li.

Academic service

2024–2026

Reviewer

NeurIPSICLRICMLKDDCIKM

Selected honors

National Scholarship for Doctoral Students · 2024, 2025
Outstanding Student, Central South University · 2025
First-class Academic Scholarship, Central South University · 2024, 2025
Microsoft Research Asia Stars of Tomorrow · Feb. 2023

Beyond research

A few personal interests

Outside research, I enjoy video creation and editing.

Favorite animation

Hyouka characters walking beneath cherry blossoms — Hyouka

Arcane ensemble artwork with the series title — Arcane

Favorite games

League of Legends gold and teal emblem — League of Legends

Hao Yan·颜浩

About me

My research interests include:

First-author research

When Graph Meets Multimodal: Benchmarking and Meditating on Multimodal Attributed Graph Learning

A Comprehensive Study on Text-attributed Graphs: Benchmarking and Rethinking

MoKGNN: Boosting Graph Neural Networks via Mixture of Generic and Task-Specific Language Models

Have Our Cake and Eat It: Augmentation Diversity and Semantic Consistency Balanced Graph Contrastive Learning

Hierarchical Graph Contrastive Learning

Rel-MOSS: Towards Imbalanced Relational Deep Learning on Relational Databases

Paradigm Shift of GNN Explainer from Label Space to Prototypical Representation Space

PS²: Parameterized Control for Fine-Grained Student Proficiency Simulation

One-Turn Knockout: Traceable and Editable Proxy Unlearning under Asymmetric Access Constraints

Unleash LLMs Potential for Sequential Recommendation by Coordinating Dual Dynamic Index Mechanism

Train Once and Explain Everywhere: Pre-training Interpretable Graph Neural Networks

WSiP: Wave Superposition Inspired Pooling for Dynamic Interactions-Aware Trajectory Prediction

Learning on Large-scale Text-attributed Graphs via Variational Inference

Adversarial Hard Negative Generation for Complementary Graph Contrastive Learning

Education

Central South University

The Hong Kong Polytechnic University

Hefei University of Technology

Research experience

Microsoft STCA

Microsoft Research Asia

Academic service

Reviewer

Selected honors

Beyond research

Favorite animation

Favorite games

Hao Yan颜浩