Erdun Gao

About Me

I am a Research Fellow at the RAIR, AIML, where I am a member of the KIPS Group led by Prof. Dino Sejdinovic. I received my Ph.D. in Applied Statistics (2025) from the School of Mathematics and Statistics at The University of Melbourne, where I was part of the Machine Learning and Reasoning (MLR) Group, advised by Prof. Mingming Gong and Prof. Howard Bondell. I also hold an M.S. in Signal and Information Processing (2020), supervised by Prof. Zhibin Pan, and a B.E. in Information Engineering (2017), both from the School of Information and Communication Engineering, Xi’an Jiaotong University.

Research Interests

My long-term research goal is to build trustworthy decision-making systems, with a focus on causal inference and uncertainty quantification. I believe that uncovering cause-and-effect relationships and properly accounting for uncertainty are key to making AI systems both robust and interpretable in high-stakes scenarios. I also occasionally work on broader machine learning problems related to data and models.

• Core Causality: Causal Discovery, Causal Effect Estimation, Causal Experimental Design.
• Causality & Machine Learning: Causal Generative Models, Causal LLMs, Causal Foundation Models.
• Machine Learning: Data Valuation, Context-Informed Learning, Uncertainty in LLMs.

Main Papers

[Full List] [Google Scholar]

• Observationally Informed Adaptive Causal Experimental Design, Preprint.
  Erdun Gao, Liang Zhang, Jake Fawkes, Aoqi Zuo, Wenqin Liu, Haoxuan Li, Mingming Gong, Dino Sejdinovic
• Causal-EPIG: A Prediction-Oriented Active Learning Framework for CATE Estimation, Preprint.
  Erdun Gao, Jake Fawkes, Dino Sejdinovic
• Concept Component Analysis: A Principled Approach for Concept Extraction in LLMs, Preprint.
  Yuhang Liu, Erdun Gao, Dong Gong, Anton van den Hengel, Javen Qinfeng Shi
• ActiveCQ: Active Estimation of Causal Quantities, ICLR, 2026.
  Erdun Gao, Dino Sejdinovic
• Domain Generalization via Content Factors Isolation: A Two-level Latent Variable Modeling Approach, MLJ, 2025.
  Erdun Gao, Howard Bondell, Shaoli Huang, Mingming Gong
• MissScore: High-Order Score Estimation in the Presence of Missing Data, ICML, 2025.
  Wenqin Liu, Haoze Hou, Erdun Gao, Biwei Huang, Qiuhong Ke, Howard Bondell, Mingming Gong
• A Variational Framework for Estimating Continuous Treatment Effects with Measurement Error, ICLR, 2024.
  Erdun Gao, Howard Bondell, Wei Huang, Mingming Gong
• Causal Discovery with Mixed Linear and Nonlinear Additive Noise Models: A Scalable Approach, CLeaR, 2024.
  Wenqin Liu, Biwei Huang, Erdun Gao, Qiuhong Ke, Howard Bondell, Mingming Gong
• FedDAG: Federated DAG Structure Learning, TMLR, 2023.
  Erdun Gao, Junjia Chen, Li Shen, Tongliang Liu, Mingming Gong, Howard Bondell
• MissDAG: Causal Discovery in the Presence of Missing Data with Continuous Additive Noise Models, NeurIPS, 2022.
  Erdun Gao, Ignavier Ng, Mingming Gong, Li Shen, Wei Huang, Tongliang Liu, Kun Zhang, Howard Bondell

Experience

• Research Fellow
  Responsible AI Research Centre, AIML & CSIRO

  Jan 2026 – Present
• Researcher
  Centre for Augmented Reasoning, AIML

  July 2024 – Dec 2025
• Postgraduate Studentship
  CSIRO

  Aug 2023 – Apr 2024
• Research Intern
  JD Explore Academy

  Mar 2021 – Mar 2022
• Research Intern
  Huawei Noah's Ark Lab

  Oct 2020 – Mar 2021

Honors & Awards

• Science Abroad Travelling Scholarships (Unimelb) 2022, 2024
• NeurIPS Scholar Award 2022
• Australian Government RTP Scholarship 2021–2024
• Outstanding Master Thesis Award (XJTU) 2020
• Outstanding Graduate (XJTU) 2017, 2020
• China National Scholarship 2019
• First-class Academic Scholarship (XJTU) 2018, 2019
• Outstanding Student (XJTU) 2014–2019
• Tangzhongying Scholarship (XJTU) 2013–2017

Academic Service

• Journal Reviewer: TMLR, TNNLS, TCSVT, MLJ, NN, JSCS
• Conference Reviewer: ICML (2022–2026), ICLR (2023–2026), NeurIPS (2023, 2025), UAI (2022–2025)
• Organising Committee: Next Generation Responsible AI Symposium (2025)