I'm a third-year PhD student in computer science supervised by Yarin Gal from OATML and Allan Dafoe from Oxford’s Centre for the Governance of AI.

My interests in machine learning include how it scales, its health applications, and safety. Before joining Oxford, I worked on learning human preferences and game-theoretical machine learning with David Duvenaud and Roger Grosse at Toronto’s Vector Institute, with the Center for Human-compatible AI at UC Berkeley, and at the Future of Humanity Institute a Oxford.

I studied machine learning (UCL), maths (Amsterdam) and Future Planet Studies (Amsterdam). I'm co-funded by Oxford and DeepMind.

Contact me

soeren.mindermann ατ gmail.com

Publications as first author

*equal contribution to first authorship
‍

Inferring the effectiveness of government interventions against COVID-19

JM Brauner*, S Mindermann*, M Sharma*, D Johnston, J Salvatier, ...
Science, 2021

Prioritized training on points that are learnable, worth learning, and not yet learned

Sören Mindermann*, Muhammed Razzak*, Winnie Xu*, Andreas Kirsch, Mrinank Sharma, Aidan Gomez, Sebastian Farquhar, Jan Brauner, Yarin Gal
ICML, 2022

Occam's razor is insufficient to infer the preferences of irrational agents

S Mindermann*, S Armstrong*
NeurIPS, 2018

Understanding the effectiveness of government interventions in Europe’s second wave of COVID-19

M Sharma*, S Mindermann*, C Rogers-Smith, G Leech, B Snodin, J Ahuja, ...
Nature Communications, 2021

Active Inverse Reward Design

S Mindermann*, R Shah*, A Gleave, D Hadfield-Menell
ICML workshop Goals in Reinforcement Learning, 2018

Identifying Causal-Effect Inference Failure with Uncertainty-Aware Models

A Jesson*, S Mindermann*, U Shalit, Y Gal
NeurIPS, 2020

Changing composition of SARS-CoV-2 lineages and rise of Delta variant in England

S Mishra*, S Mindermann*, M Sharma*, C Whittaker*, T Mellan, T Wilton, ...
The Lancet: EClinicalMedicine, 2021

How Robust are the Estimated Effects of Nonpharmaceutical Interventions against COVID-19?

M Sharma*, S Mindermann*, J Brauner*, G Leech, A Stephenson, ...
NeurIPS (Spotlight talk), 2020

Publications as senior author

*equal contribution to senior authorship
‍

Mask wearing in community settings reduces SARS-CoV-2 transmission

G Leech, C Rogers-Smith, J Sandbrink, B Snodin, R Zinkov, B Rader, J Brownstein, Y Gal, S Bhatt*, M Sharma*, S Mindermann*, J Brauner*, L Aitchison*
Proceedings of the National Academy of Sciences (PNAS), 2022

A dataset of non-pharmaceutical interventions on SARS-CoV-2 in Europe

G Altman, J Ahuja, JT Monrad, G Dhaliwal, C Rogers-Smith, G Leech, B Snodin, JB Sandbrink, L Finnveden, AJ Norman, SB Oehm, JF Sandkühler, J Kulveit, S Flaxman, Y Gal, S Mishra, S Bhatt, M Sharma*, S Mindermann*, J Brauner*
Nature Scientific Data, 2022

Publications as co-author

The Alignment Problem from a Deep Learning Perspective

R Ngo, L Chan, S Mindermann
Arxiv, 2022

Quantifying Ignorance in Individual-Level Causal-Effect Estimates under Hidden Confounding

A Jesson, S Mindermann, Y Gal, U Shalit
ICML, 2021

Is the cure really worse than the disease? The health impacts of lockdowns during COVID-19

G Meyerowitz-Katz, S Bhatt, O Ratmann, JM Brauner, S Flaxman, S Mishra, M Sharma, S Mindermann, V Bradley, M Vollmer, L Merone, G Yamey
BMJ Global Health, 2021

Seasonal variation in SARS-CoV-2 transmission in temperate climates: A Bayesian modelling study in 143 European regions

Tomáš Gavenčiak, Joshua Teperowski Monrad, Gavin Leech, Mrinank Sharma, Sören Mindermann, Jan Marus Brauner, Samir Bhatt, Jan Kulveit
PLOS Computational Biology, 2022
‍

Policy impact

Recent preprint cited in the German federal bill that decided the national lockdown in force as of May 2021
My papers on COVID-19 have been presented at the WHO, the modeling groups of the Africa CDC and the UK’s Scientific Advisory Group for Emergencies (SAGE), and the House of Representatives of the Netherlands
I presented work on mask-wearing at the UK Cabinet Office to support the UK's plan for fall 2021.

TV and newspaper interviews

Monitor TV magazine on ARD (German equiv. of BBC, ~3m viewers per episode), 2021. Talked about paper covering COVID’s 2nd wave. Interview also covered on Tagesschau.de and others.
Süddeutsche Zeitung, 2021. Interview about government interventions in COVID’s second wave.
NRC Handelsblad, 2021. Interview about government interventions in COVID’s first wave.
Turing Institute Podcast, 2021. Talked about this paper.
ITV Peston, the flagship political discussion programme of ITV, 2021 (paywalled). Talked about this paper.
DW News, 2021. Talked about this paper.

Invited talks

“Prioritized training on points that are learnable, worth learning, and not yet learned”, Meta AI Research, 2022
“Prioritized training on points that are learnable, worth learning, and not yet learned”, Cohere.ai, 2021
“Identifying Causal Effect Inference Failure with Uncertainty-Aware Models”, Vilnius Machine Learning Workshop, 2021
“Effectiveness of government interventions against COVID-19”, Africa CDC modelling consortium, 2020
“Effectiveness of government interventions against COVID-19”, German Centre for Infection Research
“Government interventions in the second wave”, ETH Zürich, 2021
“Government interventions in the second wave”, MRC Centre for Global Infectious Disease Analysis, Imperial College, 2021
“How Robust are the Estimated Effects of Nonpharmaceutical Interventions against COVID-19?”, NeurIPS Spotlight, 2020
“How Robust are the Estimated Effects of Nonpharmaceutical Interventions against COVID-19?”, NeurIPS COVID Symposium Spotlight, 2020