Sumantrak Mukherjee

Research

I work on multi-armed bandits, uncertainty quantification, and sample-efficient decision-making under limited feedback.

I am currently a Junior Researcher at DFKI in the Data Science and its Applications group, where I work under Prof. Sebastian Vollmer. I graduated from BITS Pilani in 2023 with a degree in Electronics and Instrumentation Engineering. Outside research, I like watching and discussing movies, travelling, and cooking new dishes when I am not too tired. I also keep a quiet acknowledgments page for people who have been especially kind to me.

For a more detailed account of my research, see the research page. I will add a longer research statement there over time.

Current focus

Multi-Armed Bandits and Uncertainty Quantification

Sequential decision-making with better priors, transfer, and calibrated uncertainty.

Interface

Causal Structure for Efficient Learning

Causal discovery and causal structure leveraged for sample efficiency and transfer.

Earlier work

Spatio-Temporal Event Modeling

Earlier work on interpretable point processes, benchmarks, and counterfactual event data.

selected publications

causal learning

Linear-LLM-SCM: Benchmarking LLMs for Coefficient Elicitation in Linear-Gaussian Causal Models

Kanta Yamaoka, Sumantrak Mukherjee, Thomas Gärtner, and 5 more authors

2026

Bib Paper

@misc{yamaoka2026linearllmscm,
  title = {Linear-LLM-SCM: Benchmarking LLMs for Coefficient Elicitation in Linear-Gaussian Causal Models},
  author = {Yamaoka, Kanta and Mukherjee, Sumantrak and G{\"a}rtner, Thomas and Selby, David A. and Konigorski, Stefan and H{\"u}llermeier, Eyke and Bengs, Viktor and Vollmer, Sebastian Josef},
  year = {2026},
  themes = {causal_learning},
  note = {Preprint},
}

causal learning

CLAM: Causal Spatial Disaggregation to Infer Local Effects From Coarse Data

Gerrit Großmann^*, Sumantrak Mukherjee^*, and Sebastian Josef Vollmer

* joint first authors

In NeurIPS 2025 Workshop on CauScien: Uncovering Causality in Science, 2025

Bib Paper

@inproceedings{grossmann2025clam,
  title = {CLAM: Causal Spatial Disaggregation to Infer Local Effects From Coarse Data},
  author = {Gro{\ss}mann, Gerrit and Mukherjee, Sumantrak and Vollmer, Sebastian Josef},
  booktitle = {NeurIPS 2025 Workshop on CauScien: Uncovering Causality in Science},
  year = {2025},
  themes = {causal_learning},
  note = {Workshop paper, 2025},
  author_note = {* joint first authors},
}

event modeling

SQUID: A Bayesian Approach for Physics-Informed Event Modeling

Sumantrak Mukherjee, Sebastian Vollmer, and Gerrit Grossmann

In EurIPS DiffSys Workshop, 2025

Bib Paper

@inproceedings{mukherjee2025squid,
  title = {SQUID: A Bayesian Approach for Physics-Informed Event Modeling},
  author = {Mukherjee, Sumantrak and Vollmer, Sebastian and Grossmann, Gerrit},
  booktitle = {EurIPS DiffSys Workshop},
  year = {2025},
  themes = {event_modeling},
  note = {Workshop paper, 2025},
}

bandits + uq

Co-Exploration and Co-Exploitation via Shared Structure in Multi-Task Bandits

Sumantrak Mukherjee, Serafima Lebedeva, Valentin Margraf, and 7 more authors

2025

Bib Paper

@misc{mukherjee2025multitaskbandits,
  title = {Co-Exploration and Co-Exploitation via Shared Structure in Multi-Task Bandits},
  author = {Mukherjee, Sumantrak and Lebedeva, Serafima and Margraf, Valentin and Hanselle, Jonas and Yamaoka, Kanta and Bengs, Viktor and Bhalwankar, Raj and Neigel, Peter and Hullermeier, Eyke and Vollmer, Sebastian Josef},
  year = {2025},
  themes = {bandits_uq},
  note = {Preprint},
}

event modeling

Neural Spatio Temporal Point Processes: Trends and Challenges

Sumantrak Mukherjee, Mouad Elhamdi, George Mohler, and 4 more authors

Transactions on Machine Learning Research, 2025

Bib Paper

@article{mukherjee2025nstpp,
  title = {Neural Spatio Temporal Point Processes: Trends and Challenges},
  author = {Mukherjee, Sumantrak and Elhamdi, Mouad and Mohler, George and Selby, David A. and Xie, Yao and Vollmer, Sebastian and Grossmann, Gerrit},
  journal = {Transactions on Machine Learning Research},
  year = {2025},
  themes = {event_modeling},
  note = {2025},
}

adjacent interests

X Hacking: The Threat of Misguided AutoML

Rahul Sharma, Sumantrak Mukherjee, Sergey Redyuk, and 3 more authors

In International Conference on Machine Learning, 2025

Bib Paper

@inproceedings{sharma2025xhacking,
  title = {X Hacking: The Threat of Misguided AutoML},
  author = {Sharma, Rahul and Mukherjee, Sumantrak and Redyuk, Sergey and Sipka, Andrea and Vollmer, Sebastian and Selby, David A.},
  booktitle = {International Conference on Machine Learning},
  year = {2025},
  themes = {adjacent_interests},
  note = {2025},
}

causal learning

Graph Agnostic Causal Bayesian Optimisation

Sumantrak Mukherjee^*, Mengyan Zhang^*, Seth Flaxman, and 1 more author

* joint first authors

In NeurIPS BDU Workshop, 2024

Bib Paper

@inproceedings{mukherjee2024gacbo,
  title = {Graph Agnostic Causal Bayesian Optimisation},
  author = {Mukherjee, Sumantrak and Zhang, Mengyan and Flaxman, Seth and Vollmer, Sebastian},
  booktitle = {NeurIPS BDU Workshop},
  year = {2024},
  themes = {causal_learning},
  note = {Workshop paper, 2024},
  author_note = {* joint first authors},
}

news

Mar 15, 2026	This website is now live in its first personalized version.
Dec 14, 2025	Released the preprint Co-Exploration and Co-Exploitation via Shared Structure in Multi-Task Bandits.
Dec 06, 2025	Presented CLAM: Causal Spatial Disaggregation to Infer Local Effects From Coarse Data at the NeurIPS 2025 CauScien workshop.
Oct 13, 2025	Attended the GEMINI Summer School on addressing the gender data gap in medicine and the application of N-of-1 trials, October 13-17, 2025.
Nov 05, 2024	Released the preprint Graph Agnostic Causal Bayesian Optimisation.
May 23, 2024	Visited Saarbrucken for the workshop on Safe and Causally Informed Deep RL on May 23-24, 2024.