Haiteng Zhao

Full-time researcher, Shanghai AI Lab

zhaohaiteng [AT] pku.edu.cn

Bio

I am currently a full-time researcher at Shanghai AI Lab. I obtained my Ph.D. in Machine Learning and Intelligence from Peking University in July 2025. From September 2020 to July 2025, I pursued my doctoral studies at Sigma Lab, Peking University, supervised by Prof. Zhihong Deng. I also collaborated on several exciting projects at the University of Hong Kong with Lingpeng Kong and Qi Liu. Previously, I obtained a B.Sc. degree in Psychology from Peking University.

My current focus lies in exploring whether deep learning can effectively encode human-like intelligence, encompassing transfer, generalization, reasoning and planning capacities, scientific research capabilities, and more. My work primarily spans three key areas:

  1. Machine Learning Theory: Investigating topics such as generalization theory, domain adaptation, transfer learning, and robustness. As large language models demonstrate impressive general capabilities, there is an urgent need to establish a foundational theory that can better predict these abilities.
  2. AI for Science: I am deeply interested in the potential of AI in scientific research. Scientific research represents a complete philosophical methodology that uses symbolic language to understand and interpret the world, gain new insights through experiments and reasoning, and apply those insights to create new things. It is also the fundamental driving force behind the progress of human civilization. Although today's AI has demonstrated strong general capabilities, how it can exhibit effective scientific research abilities remains an open research question.
  3. Reasoning and Planning: Addressing compositional generalization challenges in deep models, which are crucial for achieving human-like intelligence. My research in this area mainly focuses on intelligent agents.

I look forward to chatting and collaborating with you!

Publications

Most recent publications on Google Scholar.
‡ indicates equal contribution.

ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows

Qiushi Sun, Zhoumianze Liu, Chang Ma, Zichen Ding, Fangzhi Xu, Zhangyue Yin, Haiteng Zhao, Zhenyu Wu, Kanzhi Cheng, Zhaoyang Liu, Jianing Wang, Qintong Li, Xiangru Tang, Tianbao Xie, Xiachong Feng, Xiang Li, Ben Kao, Wenhai Wang, Biqing Qi, Lingpeng Kong, Zhiyong Wu

arXiv preprint

Paper

Genius: A Generalizable and Purely Unsupervised Self-Training Framework For Advanced Reasoning

Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Qiushi Sun, Kanzhi Cheng, Junxian He, Jun Liu, Zhiyong Wu

Annual Meeting of the Association for Computational Linguistics (ACL) 2025

Paper

φ-Decoding: Adaptive Foresight Sampling for Balanced Inference-Time Exploration and Exploitation

Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Jun Liu, Qika Lin, Zhiyong Wu

Annual Meeting of the Association for Computational Linguistics (ACL) 2025

Paper

BioMaze: Benchmarking and Enhancing Large Language Models for Biological Pathway Reasoning

Haiteng Zhao, Chang Ma, Fangzhi Xu, Lingpeng Kong, Zhi-Hong Deng

arXiv preprint

Paper

Non-myopic Generation of Language Model for Reasoning and Planning

Chang Ma, Haiteng Zhao, Junlei Zhang, Junxian He, Lingpeng Kong

International Conference on Learning Representations (ICLR) 2025

Paper

Instruction-Based Molecular Graph Generation with Unified Text-Graph Diffusion Model

Yuran Xiang, Haiteng Zhao, Chang Ma, Zhi-Hong Deng

arXiv preprint

Paper

Empowering Large Language Model Agents through Action Learning

Haiteng Zhao, Chang Ma, Guoyin Wang, Jing Su, Lingpeng Kong, Jingjing Xu, Zhi-Hong Deng, Hongxia Yang

Conference on Language Modeling (COLM) 2024

Paper

Exploring the Reasoning Abilities of Multimodal Large Language Models (MLLMs): A Comprehensive Survey on Emerging Trends in Multimodal Reasoning

Yiqi Wang, Wentao Chen, Xiaotian Han, Xudong Lin, Haiteng Zhao, Yongfei Liu, Bohan Zhai, Jianbo Yuan, Quanzeng You, Hongxia Yang

arXiv preprint

Paper

GIMLET: A Unified Graph-Text Model for Instruction-Based Molecule Zero-Shot Learning

Haiteng Zhao, Shengchao Liu, Chang Ma, Hannan Xu, Jie Fu, Zhi-Hong Deng, Lingpeng Kong, Qi Liu

Conference on Neural Information Processing Systems (NeurIPS) 2023

Paper

ChatPathway: Conversational Large Language Models for Biology Pathway Detection

Yanjing Li, Hannan Xu, Haiteng Zhao, Hongyu Guo, Shengchao Liu

Conference on Neural Information Processing Systems (NeurIPS) 2023 AI for Science Workshop

Paper

Are More Layers Beneficial to Graph Transformers?

Haiteng Zhao, Shuming Ma, Dongdong Zhang, Zhi-Hong Deng, Furu Wei

International Conference on Learning Representations (ICLR) 2023

Paper

Retrieved Sequence Augmentation for Protein Representation Learning

Chang Ma, Haiteng Zhao, Lin Zheng, Jiayi Xin, Qintong Li, Lijun Wu, Zhihong Deng, Yang Lu, Qi Liu, Lingpeng Kong

Conference on Empirical Methods in Natural Language Processing (EMNLP) 2024

Paper

Certified Robustness Against Natural Language Attacks by Causal Intervention

Haiteng Zhao‡, Chang Ma‡, Xinshuai Dong‡, Anh Tuan Luu, Zhi-Hong Deng, Hanwang Zhang

International Conference on Machine Learning (ICML) 2022

Paper

Domain Adaptation via Mutual Information Maximization

Haiteng Zhao, Chang Ma, Qinyu Chen, Zhihong Deng

International Joint Conference on Artificial Intelligence (IJCAI) 2022 (Long presentation)

Paper

ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows

Qiushi Sun, Zhoumianze Liu, Chang Ma, Zichen Ding, Fangzhi Xu, Zhangyue Yin, Haiteng Zhao, Zhenyu Wu, Kanzhi Cheng, Zhaoyang Liu, Jianing Wang, Qintong Li, Xiangru Tang, Tianbao Xie, Xiachong Feng, Xiang Li, Ben Kao, Wenhai Wang, Biqing Qi, Lingpeng Kong, Zhiyong Wu

arXiv preprint

Paper

BioMaze: Benchmarking and Enhancing Large Language Models for Biological Pathway Reasoning

Haiteng Zhao, Chang Ma, Fangzhi Xu, Lingpeng Kong, Zhi-Hong Deng

arXiv preprint

Paper

Instruction-Based Molecular Graph Generation with Unified Text-Graph Diffusion Model

Yuran Xiang, Haiteng Zhao, Chang Ma, Zhi-Hong Deng

arXiv preprint

Paper

GIMLET: A Unified Graph-Text Model for Instruction-Based Molecule Zero-Shot Learning

Haiteng Zhao, Shengchao Liu, Chang Ma, Hannan Xu, Jie Fu, Zhi-Hong Deng, Lingpeng Kong, Qi Liu

Conference on Neural Information Processing Systems (NeurIPS) 2023

Paper

ChatPathway: Conversational Large Language Models for Biology Pathway Detection

Yanjing Li, Hannan Xu, Haiteng Zhao, Hongyu Guo, Shengchao Liu

Conference on Neural Information Processing Systems (NeurIPS) 2023 AI for Science Workshop

Paper

Retrieved Sequence Augmentation for Protein Representation Learning

Chang Ma, Haiteng Zhao, Lin Zheng, Jiayi Xin, Qintong Li, Lijun Wu, Zhihong Deng, Yang Lu, Qi Liu, Lingpeng Kong

Conference on Empirical Methods in Natural Language Processing (EMNLP) 2024

Paper

ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows

Qiushi Sun, Zhoumianze Liu, Chang Ma, Zichen Ding, Fangzhi Xu, Zhangyue Yin, Haiteng Zhao, Zhenyu Wu, Kanzhi Cheng, Zhaoyang Liu, Jianing Wang, Qintong Li, Xiangru Tang, Tianbao Xie, Xiachong Feng, Xiang Li, Ben Kao, Wenhai Wang, Biqing Qi, Lingpeng Kong, Zhiyong Wu

arXiv preprint

Paper

Genius: A Generalizable and Purely Unsupervised Self-Training Framework For Advanced Reasoning

Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Qiushi Sun, Kanzhi Cheng, Junxian He, Jun Liu, Zhiyong Wu

Annual Meeting of the Association for Computational Linguistics (ACL) 2025

Paper

φ-Decoding: Adaptive Foresight Sampling for Balanced Inference-Time Exploration and Exploitation

Fangzhi Xu, Hang Yan, Chang Ma, Haiteng Zhao, Jun Liu, Qika Lin, Zhiyong Wu

Annual Meeting of the Association for Computational Linguistics (ACL) 2025

Paper

BioMaze: Benchmarking and Enhancing Large Language Models for Biological Pathway Reasoning

Haiteng Zhao, Chang Ma, Fangzhi Xu, Lingpeng Kong, Zhi-Hong Deng

arXiv preprint

Paper

Non-myopic Generation of Language Model for Reasoning and Planning

Chang Ma, Haiteng Zhao, Junlei Zhang, Junxian He, Lingpeng Kong

International Conference on Learning Representations (ICLR) 2025

Paper

Empowering Large Language Model Agents through Action Learning

Haiteng Zhao, Chang Ma, Guoyin Wang, Jing Su, Lingpeng Kong, Jingjing Xu, Zhi-Hong Deng, Hongxia Yang

Conference on Language Modeling (COLM) 2024

Paper

Are More Layers Beneficial to Graph Transformers?

Haiteng Zhao, Shuming Ma, Dongdong Zhang, Zhi-Hong Deng, Furu Wei

International Conference on Learning Representations (ICLR) 2023

Paper

Certified Robustness Against Natural Language Attacks by Causal Intervention

Haiteng Zhao‡, Chang Ma‡, Xinshuai Dong‡, Anh Tuan Luu, Zhi-Hong Deng, Hanwang Zhang

International Conference on Machine Learning (ICML) 2022

Paper

Domain Adaptation via Mutual Information Maximization

Haiteng Zhao, Chang Ma, Qinyu Chen, Zhihong Deng

International Joint Conference on Artificial Intelligence (IJCAI) 2022 (Long presentation)

Paper

Vitæ

Full Resume in PDF.

Thanks to Martin Saveski for the website template.
Haiteng Zhao
zhaohaiteng [AT] pku.edu.cn