Chen Xu

Chen Xu

Research Scientist, Toyota Research Institute

I work on safe and robust Large Behavior Models (i.e., generalist policies) for robot manipulation at the Toyota Research Institute (Robotics division), on the Trustworthy Learning under Uncertainty (TLU) team.

Research interests. safe & reliable generalist manipulation policies, flow-based generative modeling, uncertainty quantification.

Education. PhD in Operations Research (specializing in Machine Learning) with Prof. Yao Xie at Georgia Tech ISyE. Before that, joint BS/MS degrees with Prof. Rina Foygel Barber at the University of Chicago — MS in Statistics, BS in Computational & Applied Mathematics and Economics (data science concentration).

Email: chen.xu@tri.global  ·  Google Scholar  ·  LinkedIn

Research

Imitation Learning for Robot Manipulation

  1. Fanqi Lin, …, Chen Xu, …, Jose Barreiros. A Systematic Study of Data Modalities and Strategies for Co-training Large Behavior Models for Robot Manipulation Robotics: Science and Systems (RSS) 2026. [Website]
  2. TRI LBM division (main contributor). A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulation Science Robotics, April 2026 (Cover paper). [Website]
  3. Chen Xu, T. K. Nguyen, P. Miller, R. Lee, P. Shah, R. A. Ambrus, H. Nishimura, and M. Itkina. Can We Detect Failures Without Failure Data? Uncertainty-Aware Runtime Failure Detection for Imitation Learning Policies Robotics: Science and Systems (RSS) 2025. [Website, Poster, RSS Talk]
    (Preliminary) Uncertainty-Aware Failure Detection for Imitation Learning Robot Policies. CoRL 2024 Workshop on Safe and Robust Robot Learning for Operation in the Real World (Spotlight).

Flow-based Generative Modeling

See this introductory slide to normalizing flow and flow matching, presented at the Toyota Research Institute Machine Learning reading group.

  1. Chen Xu, Xiuyuan Cheng, and Yao Xie. Local flow matching generative models IEEE Transactions on Information Theory (TIT), 2025.
  2. Chen Xu, Jonghyeok Lee, Xiuyuan Cheng, and Yao Xie. Flow-based distributionally robust optimization IEEE Journal on Selected Areas in Information Theory (JSAIT), 2024. [Poster]
  3. Chen Xu, Xiuyuan Cheng, and Yao Xie. Normalizing flow neural networks by JKO scheme NeurIPS 2023 (Spotlight). [Poster, Slide]
  4. Chen Xu, Xiuyuan Cheng, and Yao Xie. Computing high-dimensional optimal transport by flow neural networks AISTATS 2025. [Poster]
  5. Chen Xu, Xiuyuan Cheng, and Yao Xie. Invertible neural networks for graph prediction IEEE Journal on Selected Areas in Information Theory (JSAIT), 2022. [Poster, Slide]

Conformal Prediction

See this slide summarizing my progress (up to the end of 2024) on time-series conformal prediction, with a recorded talk organized by the Time Series Analysis and Forecasting Society (TAFS).

  1. Junghwan Lee, Chen Xu, and Yao Xie. Flow-based conformal prediction for multi-dimensional time series ICLR 2026.
  2. Jonghyeok Lee, Chen Xu, and Yao Xie. Kernel-based optimally weighted conformal prediction intervals ICLR 2025.
  3. Chen Xu, Hanyang Jiang, and Yao Xie. Conformal prediction for multi-dimensional time series by ellipsoidal sets ICML 2024 (Spotlight). [Poster, Slide]
  4. Chen Xu and Yao Xie. Sequential predictive conformal inference for time series ICML 2023. [Poster, Slide]
  5. Chen Xu and Yao Xie. Conformal prediction for time series IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2023. [Poster, Slide]
    The method EnbPI has been independently implemented in six open-source libraries: MAPIE, Fortuna, skpro, TorchCP, PUNCC, ConformalPrediction.jl.
  6. Chen Xu and Yao Xie. Conformal prediction set for time-series ICML 2022 Workshop on Distribution-free Uncertainty Quantification. [Poster]
  7. Chen Xu and Yao Xie. Conformal anomaly detection on spatio-temporal observations with missing data ICML 2021 Workshop on Distribution-free Uncertainty Quantification. [Poster]
  8. Chen Xu and Yao Xie. Conformal prediction interval for dynamic time-series ICML 2021 (oral).
  9. Byol Kim, Chen Xu, and Rina Barber. Predictive inference is free with the jackknife+-after-bootstrap NeurIPS 2020.

Miscellaneous

  1. Chen Xu, Minghe Zhang, Andy Sun, Feng Qiu, and Yao Xie. Solar radiation ramping events modeling using spatio-temporal point processes INFORMS Journal on Data Science (IJDS), 2025. [Slide]
  2. Chen Xu, Yao Xie, Daniel A. Zuniga Vazquez, Rui Yao, and Feng Qiu. Spatio-temporal wildfire prediction using multi-modal data IEEE Journal on Selected Areas in Information Theory (JSAIT), 2023. [Slide]
  3. Anatoli Juditsky, Arkadi Nemirovski, Yao Xie, and Chen Xu. Generalized generalized linear models: Convex estimation and online bounds arXiv:2304.13793, 2023.
  4. Chen Xu, Xiuyuan Cheng, and Yao Xie. An alternative approach to train neural networks using monotone variational inequality OPT 2023: Optimization for Machine Learning. [Poster]