Portrait of Botao Hao

Botao Hao (郝博韬)

Research Scientist, Reasoning Team, OpenAI

I am a research scientist in the Reasoning Team at OpenAI. I work on using large-scale RL to train next-generation of reasoning models and build the smartest agent in Codex. I am a core contributor of o1 (o1 system card), o3 and GPT-5 series (GPT-5 system card). Previously, I was a Research Scientist at DeepMind and a Postdoc at Princeton University. See my talk at Stanford RL forum about information-directed sampling for explorations [Recording] [Slides].

haobotao000@gmail.com Google Scholar LinkedIn

About Me

  1. 2024-Present: Research Scientist, OpenAI
  2. 2020-2024: Research Scientist, DeepMind
  3. 2019-2020: Postdoc, Department of Electrical Engineering, Princeton University
  4. 2014-2019: Ph.D. in Statistics, Purdue University

Selected Publications

  1. GPT-5 System Card
    OpenAI, Aaditya Singh, Adam Fry, Adam Perelman, Adam Tart, Adi Ganesh, ... , Botao Hao, et al.
    arXiv 2026. arXiv
  2. OpenAI o1 System Card
    OpenAI, Aaron Jaech, Adam Kalai, Adam Lerer, Adam Richardson, Ahmed El-Kishky, ... , Botao Hao, et al.
    arXiv 2024. arXiv
  3. Leveraging Demonstrations to Improve Online Learning: Quality Matters
    Botao Hao, Rahul Jain, Tor Lattimore, Benjamin Van Roy, Zheng Wen
    ICML 2023. arXiv
  4. Regret Bounds for Information-Directed Reinforcement Learning
    Botao Hao, Tor Lattimore
    NeurIPS 2022. arXiv
  5. Contextual Information-Directed Sampling
    Botao Hao, Tor Lattimore, Chao Qin
    ICML 2022. arXiv
  6. Online Sparse Reinforcement Learning
    Botao Hao, Tor Lattimore, Csaba Szepesvari, Mengdi Wang
    AISTATS 2021. arXiv poster
  7. Information Directed Sampling for Sparse Linear Bandits
    Botao Hao, Tor Lattimore, Wei Deng
    NeurIPS 2021 (spotlight). Proceedings slides
  8. High-Dimensional Sparse Linear Bandits
    Botao Hao, Tor Lattimore, Mengdi Wang
    NeurIPS 2020. arXiv slides poster
  9. Adaptive Exploration in Linear Contextual Bandit
    Botao Hao, Tor Lattimore, Csaba Szepesvari
    AISTATS 2020. arXiv slides
  10. Sparse and Low-rank Tensor Estimation via Cubic Sketchings
    Botao Hao, Anru Zhang, Guang Cheng
    IEEE Transactions on Information Theory (2020). arXiv slides
  11. Bootstrapping Upper Confidence Bound
    Botao Hao, Yasin Abbasi-Yadkori, Zheng Wen, Guang Cheng
    NeurIPS 2019. arXiv poster
  12. Simultaneous Clustering and Estimation of Heterogeneous Graphical Models
    Botao Hao, Will Wei Sun, Yufeng Liu, Guang Cheng
    Journal of Machine Learning Research. pdf slides