Tim G. J. Rudner

Incoming Assistant Professor & Faculty Fellow, New York University

PhD Candidate, Department of Computer Science, University of Oxford

  Oxford Applied & Theoretical Machine Learning Group (OATML)

  Oxford Computational Statistics & Machine Learning Group (OxCSML)

tim.rudner [AT] cs.ox.ac.uk

     

About

I am a PhD Candidate in the Department of Computer Science at the University of Oxford, where I conduct research on probabilistic machine learning with Yarin Gal and Yee Whye Teh. The goal of my research is to develop methods and theoretical insights that enable the safe deployment of machine learning systems in safety-critical settings by drawing on tools from variational Bayesian inference and reinforcement learning. I am particularly interested in Bayesian uncertainty quantification in deep learning, probabilistic inference in reinforcement learning and control, and AI safety. For my work on safe decision-making under uncertainty, I received the 2021 Qualcomm Innovation Fellowship. I care deeply about equitable access to education and serve as an Equality, Diversity & Inclusion Fellow at the University of Oxford.

I am also an AI Fellow at Georgetown University's Center for Security and Emerging Technology and a Rhodes Scholar.

Bio

I obtained an undergraduate degree in mathematics and economics from Yale University, where I received the Charles E. Clark Memorial Award for Academic Excellence. Subsequently, I earned a master's degree in statistics from the University of Oxford, where I was advised by Dino Sejdinovic. During my PhD, I was fortunate to work with Sergey Levine at the University of California, Berkeley and with Sekhar Tatikonda at Yale University. Prior to coming to Oxford, I conducted research on game theoretic equilibria in digital goods markets, systemic risk in financial markets, and drivers of financial crises. I am a member of the Oxford Center for Doctoral Training in Autonomous Intelligent Machines & Systems (AIMS) and the OECD Network of Experts, as well as a Fellow of the German Academic Scholarship Foundation.

News

Publications

Most recent publications on Google Scholar.

Probabilistic Machine Learning



Tractable Function-Space Variational Inference in Bayesian Neural Networks

Tim G. J. Rudner*, Zonghao Chen*, Yee Whye Teh, Yarin Gal

NeurIPS '22 Conference on Neural Information Processing Systems. 2022.

  Paper      Poster    Bibtex



Continual Learning via Sequential Function-Space Variational Inference

Tim G. J. Rudner, Freddie Bickford Smith, Qixuan Feng, Yee Whye Teh, Yarin Gal

ICML '22 International Conference on Machine Learning. 2022.

  Paper     Code       Talk       Slides  Bibtex



On Signal-to-Noise Ratio Issues in Variational Inference for Deep Gaussian Processes

Tim G. J. Rudner*, Oscar Key*, Yarin Gal, Tom Rainforth

ICML '21 International Conference on Machine Learning. 2021.

   arXiv      Code       Talk       Slides     Poster    Bibtex



Inter-domain Deep Gaussian Processes

Tim G. J. Rudner, Dino Sejdinovic, Yarin Gal

ICML '20 International Conference on Machine Learning. 2020.

   arXiv    Website     Talk       Slides  Bibtex
Reinforcement Learning





Challenges and Opportunities in Offline Reinforcement Learning from Visual Observations

Cong Lu*, Philip J. Ball*, Tim G. J. Rudner, Jack Parker-Holder, Michael A. Osborne, Yee Whye Teh

Preprint. 2022.

RSS '22 Workshop on Learning from Diverse, Offline Data. 2022. (Outstanding Paper Award)

ICML '22 Workshop on Decision Awareness in Reinforcement Learning. 2022.

   arXiv      Code   Bibtex



Outcome-Driven Reinforcement Learning via Variational Inference

Tim G. J. Rudner*, Vitchyr H. Pong*, Rowan McAllister, Yarin Gal, Sergey Levine

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper   Website     Talk       Slides     Poster    Bibtex



On Pathologies in KL-Regularized Reinforcement Learning from Expert Demonstrations

Tim G. J. Rudner*, Cong Lu*, Michael A. Osborne, Yarin Gal, Yee Whye Teh

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper     Code   Website     Talk       Slides     Poster    Bibtex



VIREL: A Variational Inference Framework for Reinforcement Learning

Matthew Fellows, Anuj Mahajan, Tim G. J. Rudner, Shimon Whiteson

NeurIPS '19 Conference on Neural Information Processing Systems. 2019. (Spotlight Talk)

   arXiv      Code   Bibtex
Reliable Uncertainty Quantification in Deep Learning










Plex: Towards Reliability Using Pretrained Large Model Extensions

Dustin Tran, Jeremiah Liu, Michael W. Dusenberry, Du Phan, Mark Collier,
Jie Ren, Kehang Han, Zi Wang, Zelda Mariet, Huiyi Hu, Neil Band,
Tim G. J. Rudner, Karan Singhal, Zachary Nado, Joost van Amersfoort,
Andreas Kirsch, Rodolphe Jenatton, Nithum Thain, Honglin Yuan,
Kelly Buchanan, Kevin Murphy, D. Sculley, Yarin Gal, Zoubin Ghahramani,
Jasper Snoek, Balaji Lakshminarayanan

Preprint. 2022.

ICML '22 Workshop on Pre-training: Perspectives, Pitfalls, and Paths Forward. 2022. (Contributed Talk)

ICML '22 Workshop on Principles of Distribution Shift. 2022.

   arXiv      Code       Talk       Blog   Bibtex








Uncertainty Baselines: Benchmarks for Uncertainty & Robustness in Deep Learning

Zachary Nado*, Neil Band*, Mark Collier, Josip Djolonga, Michael W. Dusenberry,
Sebastian Farquhar, Qixuan Feng, Angelos Filos, Marton Havasi, Rodolphe Jenatton,
Ghassen Jerfel, Jeremiah Liu, Zelda Mariet, Jeremy Nixon, Shreyas Padhy, Jie Ren,
Tim G. J. Rudner, Faris Sbahi, Yeming Wen, Florian Wenzel, Kevin Murphy,
D. Sculley, Balaji Lakshminarayanan, Jasper Snoek, Yarin Gal, Dustin Tran

Technical Report. 2022.

NeurIPS '21 Workshop on Bayesian Deep Learning. 2021.

   arXiv      Code     Blog   Bibtex




Benchmarking Bayesian Deep Learning on Diabetic Retinopathy Detection Tasks

Neil Band*, Tim G. J. Rudner*, Qixuan Feng, Angelos Filos, Zachary Nado,
Michael W. Dusenberry, Ghassen Jerfel, Dustin Tran, Yarin Gal

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper     Code      Poster    Bibtex
Non-Technical Papers

Key Concepts in AI Safety: Specification in Machine Learning

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper    Media  Bibtex

Key Concepts in AI Safety: Interpretability in Machine Learning

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

Key Concepts in AI Safety: Robustness and Adversarial Examples

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

Key Concepts in AI Safety: An Overview

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

OECD Framework for the Classification of AI systems

OECD (contributing author: Tim G. J. Rudner)

OECD Publishing. 2022.

  Paper   Bibtex
Preprints





Challenges and Opportunities in Offline Reinforcement Learning from Visual Observations

Cong Lu*, Philip J. Ball*, Tim G. J. Rudner, Jack Parker-Holder, Michael A. Osborne, Yee Whye Teh

Preprint. 2022.

RSS '22 Workshop on Learning from Diverse, Offline Data. 2022. (Outstanding Paper Award)

ICML '22 Workshop on Decision Awareness in Reinforcement Learning. 2022.

   arXiv      Code   Bibtex










Plex: Towards Reliability Using Pretrained Large Model Extensions

Dustin Tran, Jeremiah Liu, Michael W. Dusenberry, Du Phan, Mark Collier,
Jie Ren, Kehang Han, Zi Wang, Zelda Mariet, Huiyi Hu, Neil Band,
Tim G. J. Rudner, Karan Singhal, Zachary Nado, Joost van Amersfoort,
Andreas Kirsch, Rodolphe Jenatton, Nithum Thain, Honglin Yuan,
Kelly Buchanan, Kevin Murphy, D. Sculley, Yarin Gal, Zoubin Ghahramani,
Jasper Snoek, Balaji Lakshminarayanan

Preprint. 2022.

ICML '22 Workshop on Pre-training: Perspectives, Pitfalls, and Paths Forward. 2022. (Contributed Talk)

ICML '22 Workshop on Principles of Distribution Shift. 2022.

   arXiv      Code       Talk       Blog   Bibtex
2022



Tractable Function-Space Variational Inference in Bayesian Neural Networks

Tim G. J. Rudner*, Zonghao Chen*, Yee Whye Teh, Yarin Gal

NeurIPS '22 Conference on Neural Information Processing Systems. 2022.

  Paper   Bibtex



Continual Learning via Sequential Function-Space Variational Inference

Tim G. J. Rudner, Freddie Bickford Smith, Qixuan Feng, Yee Whye Teh, Yarin Gal

ICML '22 International Conference on Machine Learning. 2022.

  Paper     Code       Talk       Slides  Bibtex








Uncertainty Baselines: Benchmarks for Uncertainty & Robustness in Deep Learning

Zachary Nado*, Neil Band*, Mark Collier, Josip Djolonga, Michael W. Dusenberry,
Sebastian Farquhar, Qixuan Feng, Angelos Filos, Marton Havasi, Rodolphe Jenatton,
Ghassen Jerfel, Jeremiah Liu, Zelda Mariet, Jeremy Nixon, Shreyas Padhy, Jie Ren,
Tim G. J. Rudner, Faris Sbahi, Yeming Wen, Florian Wenzel, Kevin Murphy,
D. Sculley, Balaji Lakshminarayanan, Jasper Snoek, Yarin Gal, Dustin Tran

Technical Report. 2022.

NeurIPS '21 Workshop on Bayesian Deep Learning. 2021.

   arXiv      Code     Blog   Bibtex
2021



Outcome-Driven Reinforcement Learning via Variational Inference

Tim G. J. Rudner*, Vitchyr H. Pong*, Rowan McAllister, Yarin Gal, Sergey Levine

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper   Website     Talk       Slides     Poster    Bibtex



On Pathologies in KL-Regularized Reinforcement Learning from Expert Demonstrations

Tim G. J. Rudner*, Cong Lu*, Michael A. Osborne, Yarin Gal, Yee Whye Teh

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper     Code   Website     Talk       Slides     Poster    Bibtex




Benchmarking Bayesian Deep Learning on Diabetic Retinopathy Detection Tasks

Neil Band*, Tim G. J. Rudner*, Qixuan Feng, Angelos Filos, Zachary Nado,
Michael W. Dusenberry, Ghassen Jerfel, Dustin Tran, Yarin Gal

NeurIPS '21 Conference on Neural Information Processing Systems. 2021.

  Paper     Code      Poster    Bibtex



On Signal-to-Noise Ratio Issues in Variational Inference for Deep Gaussian Processes

Tim G. J. Rudner*, Oscar Key*, Yarin Gal, Tom Rainforth

ICML '21 International Conference on Machine Learning. 2021.

   arXiv      Code       Talk       Slides     Poster    Bibtex
2020



Inter-domain Deep Gaussian Processes

Tim G. J. Rudner, Dino Sejdinovic, Yarin Gal

ICML '20 International Conference on Machine Learning. 2020.

   arXiv    Website     Talk       Slides  Bibtex
2019



VIREL: A Variational Inference Framework for Reinforcement Learning

Matthew Fellows, Anuj Mahajan, Tim G. J. Rudner, Shimon Whiteson

NeurIPS '19 Conference on Neural Information Processing Systems. 2019. (Spotlight Talk)

   arXiv      Code   Bibtex



The Natural Neural Tangent Kernel: Neural Network Training Dynamics under Natural Gradient Descent

Tim G. J. Rudner, Florian Wenzel, Yee Whye Teh, Yarin Gal

NeurIPS '19 Workshop on Bayesian Deep Learning. 2019. (Contributed Talk)

  Paper       Talk     Bibtex





A Systematic Comparison of Bayesian Deep Learning Robustness in Diabetic Retinopathy Tasks

Angelos Filos, Sebastian Farquhar, Aidan N. Gomez, Tim G. J. Rudner, Zachary Kenton, Lewis Smith, Milad Alizadeh,
Arnoud de Kroon, Yarin Gal

Technical Report. 2019.

NeurIPS '19 Workshop on Bayesian Deep Learning. 2019.

   arXiv      Code   Bibtex



Multi3Net: Segmenting Flooded Buildings via Fusion of Multiresolution, Multisensor, and Multitemporal Satellite Imagery

Tim G. J. Rudner, Marc Rußwurm, Jakub Fil, Ramona Pelich, Benjamin Bischke, Veronika Kopackova, Piotr Bilinski

AAAI '19 Conference on Artificial Intelligence. 2019.

   arXiv      Code    Media  Bibtex




The StarCraft Multi-Agent Challenge

Mikayel Samvelyan, Tabish Rashid, Christian Schroeder de Witt, Gregory Farquhar, Nantas Nardelli,
Tim G. J. Rudner, Chia-Man Hung, Philip H. S. Torr, Jakob Foerster, Shimon Whiteson

AAMAS '19 International Conference on Autonomous Agents and Multiagent Systems. 2019.

   arXiv      Code   Bibtex
2018



On the Connection between Neural Processes and Gaussian Processes with Deep Kernels

Tim G. J. Rudner, Vincent Fortuin, Yee Whye Teh, Yarin Gal

NeurIPS '18 Workshop on Bayesian Deep Learning. 2018.

  Paper   Bibtex
Non-Technical Papers

Key Concepts in AI Safety: Specification in Machine Learning

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper    Media  Bibtex

Key Concepts in AI Safety: Interpretability in Machine Learning

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

Key Concepts in AI Safety: Robustness and Adversarial Examples

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

Key Concepts in AI Safety: An Overview

Tim G. J. Rudner and Helen Toner

CSET Issue Briefs. 2021.

  Paper   Bibtex

OECD Framework for the Classification of AI systems

OECD (contributing author: Tim G. J. Rudner)

OECD Publishing. 2022.

  Paper   Bibtex

Teaching (Past & Present)