Hao Liu

Email: haoliu4221@gmail.com

Seattle, WA

I am an Applied Scientist at Amazon Web Services (AWS), where I work on agentic AI systems and post-training for tool-using large language models.

My recent work focuses on autonomous agents for software and operational workflows, including agent orchestration, context gathering, evaluation-driven system improvement, and SFT/RL-based training. I am especially interested in the intersection of agent scaffolding and model improvement: building systems that can plan, use tools, interact with external environments, and improve through post-training.

Before industry, I completed my Ph.D. in Computer Science at Washington University in St. Louis, advised by Professor Yixin Chen. My research focused on foundation models for structured data, with work spanning unified modeling across tasks, self-supervised pretraining, few-shot and zero-shot adaptation, and the integration of language models with structure-aware neural architectures. My research has been published at ICLR, NeurIPS, and WWW.

Previously, I worked on graph foundation models, LLM-GNN systems, few-shot learning, and tabular representation learning. While my current work is centered on agentic AI, I continue to be interested in model generalization, scalable training, and research that bridges strong technical ideas with real-world systems.

You can find more details here:

selected works

GOFA: A Generative One-For-All Model for Joint Graph Language Modeling

*Lecheng Kong, *Jiarui Feng, *Hao Liu, and 4 more authors

In The Thirteenth International Conference on Learning Representations, 2025

Bib PDF Code

@inproceedings{kong2024gofa,
  title = {GOFA: A Generative One-For-All Model for Joint Graph Language Modeling},
  author = {Kong, *Lecheng and Feng, *Jiarui and Liu, *Hao and Huang, Chengsong and Huang, Jiaxin and Chen, Yixin and Zhang, Muhan},
  booktitle = {The Thirteenth International Conference on Learning Representations},
  year = {2025},
}

TAGLAS: An atlas of text-attributed graph datasets in the era of large graph and language models

Jiarui Feng, *Hao Liu, *Lecheng Kong, and 2 more authors

arXiv preprint arXiv:2406.14683, 2024

Bib PDF Code

@article{feng2024taglas,
  title = {TAGLAS: An atlas of text-attributed graph datasets in the era of large graph and language models},
  author = {Feng, Jiarui and Liu, *Hao and Kong, *Lecheng and Chen, Yixin and Zhang, Muhan},
  journal = {arXiv preprint arXiv:2406.14683},
  year = {2024},
}

One for All: Towards Training One Graph Model for All Classification Tasks

*Hao Liu, *Jiarui Feng, *Lecheng Kong, and 4 more authors

In The Twelfth International Conference on Learning Representations, 2024

Bib PDF Code Slides

@inproceedings{liu2023one,
  title = {One for All: Towards Training One Graph Model for All Classification Tasks},
  author = {Liu, *Hao and Feng, *Jiarui and Kong, *Lecheng and Liang, Ningyue and Tao, Dacheng and Chen, Yixin and Zhang, Muhan},
  booktitle = {The Twelfth International Conference on Learning Representations},
  year = {2024},
}

Graph Contrastive Learning Meets Graph Meta Learning: A Unified Method for Few-shot Node Tasks

Hao Liu, Jiarui Feng, Lecheng Kong, and 3 more authors

In The Web Conference 2024, 2024

Bib PDF Code

@inproceedings{liu2023graph,
  title = {Graph Contrastive Learning Meets Graph Meta Learning: A Unified Method for Few-shot Node Tasks},
  author = {Liu, Hao and Feng, Jiarui and Kong, Lecheng and Tao, Dacheng and Chen, Yixin and Zhang, Muhan},
  booktitle = {The Web Conference 2024},
  year = {2024},
}

NeurIPS TRL 2023

TabContrast: A Local-Global Level Method for Tabular Contrastive Learning

Hao Liu, Yixin Chen, Bradley Fritz, and 1 more author

In NeurIPS 2023 Second Table Representation Learning Workshop, 2023

Bib

@inproceedings{liu2023tabcontrast,
  title = {TabContrast: A Local-Global Level Method for Tabular Contrastive Learning},
  author = {Liu, Hao and Chen, Yixin and Fritz, Bradley and King, Christopher},
  booktitle = {NeurIPS 2023 Second Table Representation Learning Workshop},
  year = {2023},
}

NeurIPS 2023

MAG-GNN: Reinforcement Learning Boosted Graph Neural Network

Lecheng Kong, Jiarui Feng, Hao Liu, and 3 more authors

In Thirty-seventh Conference on Neural Information Processing Systems, 2023

Bib

@inproceedings{kong2023mag,
  title = {MAG-GNN: Reinforcement Learning Boosted Graph Neural Network},
  author = {Kong, Lecheng and Feng, Jiarui and Liu, Hao and Tao, Dacheng and Chen, Yixin and Zhang, Muhan},
  booktitle = {Thirty-seventh Conference on Neural Information Processing Systems},
  year = {2023},
}

NeurIPS 2023

Extending the Design Space of Graph Neural Networks by Rethinking Folklore Weisfeiler-Lehman

Jiarui Feng, Lecheng Kong, Hao Liu, and 4 more authors

In Thirty-seventh Conference on Neural Information Processing Systems, 2023

Bib

@inproceedings{feng2023extending,
  title = {Extending the Design Space of Graph Neural Networks by Rethinking Folklore Weisfeiler-Lehman},
  author = {Feng, Jiarui and Kong, Lecheng and Liu, Hao and Tao, Dacheng and Li, Fuhai and Zhang, Muhan and Chen, Yixin},
  booktitle = {Thirty-seventh Conference on Neural Information Processing Systems},
  year = {2023},
}

education

2019.09 - 2025.05, Ph.D., Computer Science and Engineering, Washington University in St. Louis.
2015.09 - 2019.06, Bachelor, Mathematics, Beijing Normal University.

personal

I'm from Jilin, China.
🎮 Roguelike Game Lover.