Yuxiao Cheng

I am Yuxiao Cheng, a Ph.D. student in the Department of Automation, Tsinghua University, advised by Prof. Jinli Suo.

My research focuses on AI for Healthcare, specifically on causal modelling, AI Agents, and multi-modal learning in digital healthcare.

I received my B.E. in Department of Automation from Tsinghua University in 2022 and am currently pursuing my Ph.D. in the Department of Automation, advised by Prof. Jinli Suo. I won the National Scholarship (国家奖学金) in 2024.

I have published 9 first/co-first author papers in top-tier venues including The Lancet Digital Health, Nature Biomedical Engineering, Nature Communications, PNAS, ICLR, and AAAI.

🔥 News

• 2025.11 🎉 One paper accepted to AAAI 2026, including “COGS: A Causal Representation Learning Framework for Out-of-Distribution Generalization in Time Series”.
• 2025.09 🎉 My project "OpenLens AI: Fully Autonomous Research Agent for Health Infomatics" is launched and prompted by medias including 量子位.
• 2025.09 🎉 Our paper “Causal deep learning for real-time detection of cardiac surgery-associated acute kidney injury” is published by The Lancet Digital Health.
• 2025.07 🎉 Our paper “A generative model uses healthy and diseased image pairs for pixel-level chest X-ray pathology localization” is published in Nature Biomedical Engineering.

🎖 Honors and Awards

• 2024.10 🏆 National Scholarship (国家奖学金), Tsinghua University

📝 First/co-first Author Papers

(* indicates equal contribution)

AI Agents for Healthcare

Open Source Project (2025-2026)

OpenLens AI: Fully Autonomous Multimodal Agent for Health Informatics Research

Yuxiao Cheng, Jinli Suo

A fully autonomous multimodal agent designed for medical/ML/stats research, or any data-driven project, and is optimized for medical + AI research.

[Paper] [Code 150+ Star] [App] [量子位]

Causal Learning in Healthcare

The Lancet Digital Health (2025)

Causal deep learning for real-time detection of cardiac surgery-associated acute kidney injury: derivation and validation in seven time-series cohorts

Qin Zhong*, Yuxiao Cheng*, Zongren Li*, Dongjin Wang*, Chongyou Rao, et al.
(I am the lead contributor for AI algorithm.)

A causal deep learning approach that combines neural networks with causal discovery to develop a reliable and generalizable model to predict a patient's risk of developing CSA-AKI.

[Paper] [Code 100+ Star] [Public Mention]

ICLR (2024)

CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discovery

Yuxiao Cheng*, Ziqian Wang*, Tingxiong Xiao, Qin Zhong, Jinli Suo, Kunlun He

[Paper] [Website] [Code] [Dataset]

A novel pipeline capable of generating realistic time-series along with a ground truth causal graph that is generalizable to different fields.

AAAI Oral (2026)

COGS: A Causal Representation Learning Framework for Out-of-Distribution Generalization in Time Series

Xinxin Song*, Yuxiao Cheng*, Tingxiong Xiao, Jinli Suo

[Paper] [Code]

A novel framework that incorporates causal representation learning into the OOD generalization of time series.

AAAI (2024)

CUTS+: High-dimensional Causal Discovery from Irregular Time-series

Yuxiao Cheng*, Lianglong Li*, Tingxiong Xiao, Zongren Li, Jinli Suo, Kunlun He, Qionghai Dai

[Paper] [Code]

Increasing scalability of neural causal discovery on high-dimensional irregular data.

ICLR (2023)

CUTS: Neural Causal Discovery from Irregular Time-Series Data

Yuxiao Cheng*, Runzhao Yang*, Tingxiong Xiao, Zongren Li, Jinli Suo, Kunlun He, Qionghai Dai

[Paper] [Code]

EM-Style joint causal graph learning and missing data imputation for irregular temporal data.

Multimodal AI in Biomedicine

Nature Biomedical Engineering (2025)

A generative model uses healthy and diseased image pairs for pixel-level chest X-ray pathology localization

Kaiming Dong*, Yuxiao Cheng*, Kunlun He, Jinli Suo

A generative model that leverages paired healthy–diseased X-rays for interpretable pathology localization.

[Paper]

Information Fusion (2023)

A mutually boosting dual sensor computational camera for high quality dark videography

Yuxiao Cheng, Runzhao Yang, Zhihong Zhang, Jinli Suo, Qionghai Dai

A supervised deep-learning denoising method that enables single-molecule FRET with up to 10-fold reduction in photon requirement per frame.

[Paper]

Nature Communications (2025)

Supervised multi-frame dual-channel denoising enables long-term single-molecule FRET under extremely low photon budget

Yu Miao*, Yuxiao Cheng*, Yushi Xia, Yongzhen Hei, et al.
(I am the lead contributor for AI algorithm.)

A supervised deep-learning denoising method that enables single-molecule FRET with up to 10-fold reduction in photon requirement per frame.

[Paper]

PNAS (2024)

Sharing massive biomedical data at magnitudes lower bandwidth using implicit neural function

Runzhao Yang*, Tingxiong Xiao*, Yuxiao Cheng*, Anan Li*, et al.

Represent the high-dimensional biomedical data volume with a compact implicit neural function and successfully reduce the demanding bandwidth by 2-3 orders of magnitude at high data fidelity

[Paper]

🎓 Education

• 2022.09 – Present Ph.D. in Automation, Tsinghua University
• 2018.09 – 2022.07 B.Eng. in Automation, Tsinghua University

💻 Internships and Research Visits

• 2026.01 – 06 (Planned) Visiting Researcher at KTH Royal Institute of Technology, Stockholm, Sweden.
• 2024.07 – 08 Research Intern, Suzhou Automotive Research Institute, real-time detection algorithms with LiDAR.
• 2021.07 – 09 Product Manager Intern, Xiaomi Inc., photo deblurring based on deep learning.

💬 Academic Services

Reviewer for ICLR 2025, Pattern Recognition, IEEE Internet of Things Journal.