Haotian Xue 薛昊天

I am a first-year Ph.D. student at ML@GaTech, advised by Prof. Yongxin Chen. Previously, I obtained my B.E. of Computer Science from Shanghai Jiao Tong University with honor in 2022.

  • [2022.12] I am selected as the Top Reviewer of NeurIPS 2022.
  • [2022.10] One co-first-author paper is accepted to EMNLP2022 Findings.
  • [2022.08] I will start as a PhD student at (ML@GT) starting from 2022Fall.
Research Interest

My research interest lies in broad aspects of Machine Learning, Computer Vision and Natural Language Processing. Currenty, I target myself to the following directions:

  • Generative models + X: utilize/learn strong prior knowledge using generative modeling, to faciliate AI problem, including robust AI, robot learning and inverse problem
  • Compositional and Explainable AI: learning Compositional and Explainable representation for deep learning models, Computer Vision and Natural Language Processing
Research Experience
  • [2023-current]: Start collaborating with Prof. Animesh Garg, GaTech, on DM+RL
  • [2023-current]: Start collaborating with Prof. Bin Hu, UIUC, on DM+Robustness
  • [2022-current]: Work as a GRA Ph.D. student at FLAIR lab with Prof. Yongxin Chen
  • [2022-2023]: Work as a remote intern at MIT CSAIL, advised by Josh Tenenbaum, Yunzhu Li and Fish Tung
  • [2021-2021]: Work as a research intern at NLC group, Microsoft Research
  • [2021-2022]: Work as a research intern at John Hopcroft Center, advised by Prof. Zhouhan Lin on NLP
  • [2020-2021]: Work as a research intern at John Hopcroft Center, advised by Prof. Quanshi Zhang on XAI
Reviewer Experience

ICML'22, NeurIPS'22, ICML'23

Publications ( show selected / show all by date / show all by topic )

Topics: Explainable A / Vision / NLP (*/†: indicates equal contribution.)

3D-IntPhys: Learning 3D Visual Intuitive Physics for Fluids, Rigid Bodies, and Granular Materials
Haotian Xue, Antonio Torralba, Joshua B. Tenenbaum, Daniel LK Yamins, Yunzhu Li, Hsiao-Yu Tung

[In Submission]

Syntax-guided Localized Self-attention by Constituency Syntactic Distance
Shengyuan Hou*, Haotian Xue*, Jushi Kai*, Bingyu Zhu, Bo Yuan, Longtao Huang, Xinbin Wang, Zhouhan Lin

[EMNLP2022, Findings]

Learning to Adaptively Incorporate External Syntax through Gated Self-Attention

[In Submission to ACL2023]

A Hypothesis For The Cognitive Difficulty of Images
Xu Chen*, Xin Wang*, Haotian Xue, Zhengyang Liang, Xin Jin, Quanshi Zhang,


Evaluation of Attribution Explanations without Ground Truth
Hao Zhang, Haotian Xue, Jiayi Chen, Yiting Chen, Wen Shen, Quanshi Zhang,