Jianguo Huang

Jianguo Huang
黄建国

My research interests include Robust Deep Learning, Uncertainty Quantification (esp., Conformal Prediction), Computer Vision.

I am currently a research associate in the AMI Research Group led by Prof. Bo An. Previously, I was a research intern in Prof. Hongxin Wei’s research group at the Southern University of Science and Technology. I completed my M.S. in Computer Science and Technology at ShanghaiTech University in 2024 under the supervision of Prof. Yue Qiu and Prof. Hao Wang, and I earned my B.S. in Information and Computation Science from North China Electric Power University in 2021.

✡️ Research Interests

My current research interests mainly focus on complex reasoning of LLM, Conformal Prediction, Uncertainty Quantification and AI4Science.

📝 Publications & Preprints

(* Corresponding author; † Equal contribution)

  • PAC Reasoning: Controlling the Performance Loss for Efficient Reasonin, Hao Zeng†, Jianguo Huang†, Bingyi Jing, Hongxin Wei, Bo An*. [paper].
  • TorchCP: A Python Library for Conformal Prediction, Jianguo Huang, Jianqing Song, Jianqing Song, Xuanning Zhou, Bingyi Jing, Hongxin Wei*. [paper], [code]. JMLR.
  • Let’s Revise Step-by-Step: A Unified Local Search Framework for Code Generation with LLMs, Zhiyi Lyu, Jianguo Huang, Yanchen Deng, Steven Hoi, Bo An*. [paper]. NeruIPS’2025.
  • Conformal Prediction Meets Long-tail Classification, Shuqi Liu, Jianguo Huang, Luke Ong. [paper]. AAAI’2026.
  • Conformal Prediction for Deep Classifier via Label Ranking. Jianguo Huang†, Huajun Xi†, Linjun Zhang, Huaxiu Yao, Yue Qiu*, Hongxin Wei*. [paper], [code]. ICML’2024.
  • Does Confidence Calibration Help Conformal Prediction? Huajun Xi†, Jianguo Huang†, Lei Feng, Hongxin Wei*. [paper]. TMLR.
  • Similarity-Navigated Conformal Prediction for Graph Neural Networks. Jianqing Song, Jianguo Huang, Wenyu Jiang, Baoming Zhang , Shuangjie Li, Chongjun Wang*. [paper]. NeurIPS’2024.
  • Resolution invariant deep operator network for PDEs with complex geometries. Jianguo Huang, Yue Qiu*. [paper], [code], [video]. Journal of Computational Physics.
  • Koopman operator learning using invertible neural networks. Yuhuang Meng†, Jianguo Huang†, Yue Qiu*. [paper], [video]. Journal of Computational Physics.

🗡️ Projects

sym

A curated and continuously updated list tracking the latest developments in Diffusion Language Models.

sym

An open-source library focused on sampling algorithms for Diffusion Language Models (DLMs).

sym

Python toolbox for conformal prediction research on deep learning models, using PyTorch. The technical report can give you a quickstart.

sym

A tutorial on optimization methods in deep learning.

📖 Educations

  • 2021.09 - 2024.06, M.S., ShanghaiTech University, Shanghai, China.
  • 2017.09 - 2021.06, B.S., North China Electric Power University, Baoding, China.

💻 Work Experience

  • Research Intern (2023.07 - 2024-08):
    • Department of Statistics and Data Science, Southern University of Science and Technology.
    • Supervisor: Prof. Hongxin Wei
  • Co-founder & Algorithm Engineer (2021.09 - Current):
    • Changsha Hongcheng Electric Technology Co., LTD.
    • Vision statement: To empower partial discharge detection with AI

🎖️ Awards

Nov. 2024. I am honored to be recognized as Top Reviewers in NeurIPS 2024.

🧑‍🏫 Teaching Assistant

  • Course CS182 - Introduction of Machine Learning, ShanghaiTech University, 2023 Spring.
  • Course TC201 - Matrix Computation, ShanghaiTech University, 2022 Autumn