Marcel Hedman

Jesus College,

Turl St

Oxford, OX1 3DW

I am a DPhil candidate in Statistical Machine Learning at the University of Oxford, supported by the StatML fellowship and supervised by Professor Chris Holmes and Professor Tom Rainforth. My research focuses on sequential decision making, with particular interests in multi-agent reinforcement learning, world modelling and Bayesian experimental design.

Previously, I read Natural Sciences (Physics) at the University of Cambridge and was a Choate Memorial Fellow at Harvard University, where I worked on data science at the intersection of machine learning and neurodegenerative disease. Before returning to academia, I was in the founding team at Axion Ray, applying AI to advanced industries, where amongst other things, I built the initial product-lead team and led deployments across enterprise accounts. Outside of research, I run Nural, an early-stage AI fund.

If you’d like to get in touch, feel free to email me at: <first_name>.<last_name>@jesus.ox.ac.uk or for startup and investment related queries at <first_name>@nural.cc

news

Jul 10, 2026	Molten Pot won Best Paper Award Runner-Up @ ICML 2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation!
Jun 24, 2026	I am presenting FairBED at Manchester Workshop on Bayesian Experimental Design 2026!
May 23, 2026	CODA was accepted at ICML 2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation!
May 23, 2026	Molten Pot was accepted at ICML 2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation!
May 05, 2026	CODA was accepted at RLC 2026!

selected publications

ARXIV

FairBED: A Bayesian Experimental Design Approach to Gathering Fairer Data

Marcel Hedman, Emily Alger, Brieuc Lehmann, and 2 more authors

2026

Bib

@misc{hedman2026fairbedbayesianexperimentaldesign,
  title = {FairBED: A Bayesian Experimental Design Approach to Gathering Fairer Data},
  author = {Hedman, Marcel and Alger, Emily and Lehmann, Brieuc and Holmes, Chris and Rainforth, Tom},
  year = {2026},
  eprint = {2606.23515},
  archiveprefix = {arXiv},
  primaryclass = {stat.ML},
  url = {https://arxiv.org/abs/2606.23515},
}

ICML Workshop

Molten Pot: Evaluations & Datasets for Social Offline Reinforcement Learning

Juan Claude Formanek^*, Marcel Hedman^*, Kale-ab Tessera, and 2 more authors

In Decision-Making from Offline Datasets to Online Adaptation: Black-Box Optimization to Reinforcement Learning, 2026

Bib

@inproceedings{formanek2026moltenpot,
  title = {Molten Pot: Evaluations \& Datasets for Social Offline Reinforcement Learning},
  author = {Formanek, Juan Claude and Hedman, Marcel and Tessera, Kale-ab and Holmes, Christopher C. and Shock, Jonathan P.},
  booktitle = {Decision-Making from Offline Datasets to Online Adaptation: Black-Box Optimization to Reinforcement Learning},
  year = {2026},
  url = {https://openreview.net/forum?id=0NYZ3Hkamd},
}

RLC

CODA: Coordination via On-Policy Diffusion for Multi-Agent Offline Reinforcement Learning

Marcel Hedman, Kale-ab Abebe Tessera, Juan Claude Formanek, and 5 more authors

2026

Bib

@misc{hedman2026codacoordinationonpolicydiffusion,
  title = {CODA: Coordination via On-Policy Diffusion for Multi-Agent Offline Reinforcement Learning},
  author = {Hedman, Marcel and Tessera, Kale-ab Abebe and Formanek, Juan Claude and Sims, Anya and Zamboni, Riccardo and McInroe, Trevor and Torr, John and Fosong, Elliot},
  year = {2026},
  eprint = {2604.23308},
  archiveprefix = {arXiv},
  primaryclass = {cs.LG},
  url = {https://arxiv.org/abs/2604.23308},
}

ICML

Step-DAD: semi-amortized policy-based Bayesian experimental design

Marcel Hedman^*, Desi R. Ivanova^*, Cong. Guan, and 1 more author

In Proceedings of the 42nd International Conference on Machine Learning, 2025

Bib Code

@inproceedings{hedman2025a,
  title = {Step-DAD: semi-amortized policy-based Bayesian experimental design},
  author = {Hedman, Marcel and Ivanova, Desi R. and Guan, Cong. and Rainforth, Tom. W. G.},
  booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
  series = {Proceedings of Machine Learning Research},
  publisher = {PMLR},
  year = {2025},
  url = {https://icml.cc/virtual/2025/poster/45693},
}