Nearchos Potamitis

Department of Computer Science

Aarhus University

Aarhus, Denmark

I am a PhD student at the Department of Computer Science at Aarhus University, advised by Prof. Akhil Arora. My research focuses on efficiency and reliability in LLM reasoning.

Previously, I obtained my M.Sc. in Data Science and B.Sc. in Electrical Engineering from EPFL. During my master’s thesis at DLAB, I introduced Fleet of Agents, a framework utilizing LLMs as agents for coordinated problem solving, receiving a perfect grade of 6/6 (top 1%).

I have also worked as a Schuman Trainee at the AI unit of the European Parliament and as a Data Science Intern at Prediggo SA.

selected publications

ICML

Fleet of Agents: Coordinated Problem Solving with Large Language Models

Lars Klein^*, Nearchos Potamitis^*, Roland Aydin, and 3 more authors

In International Conference on Machine Learning (ICML), 2025

Bib

@inproceedings{klein2025fleet,
  title = {Fleet of Agents: Coordinated Problem Solving with Large Language Models},
  author = {Klein, Lars and Potamitis, Nearchos and Aydin, Roland and Robert, West and Caglar, Gulcehre and Arora, Akhil},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2025},
}

EMNLP

CacheSaver: A Modular Framework for Efficient, Affordable, and Reproducible LLM Inference

Nearchos Potamitis^*, Lars Klein^*, Bardia Mohammadi, and 5 more authors

In Findings of the Association for Computational Linguistics: EMNLP, 2025

Bib

@inproceedings{potamitis2025cachesaver,
  title = {CacheSaver: A Modular Framework for Efficient, Affordable, and Reproducible LLM Inference},
  author = {Potamitis, Nearchos and Klein, Lars and Mohammadi, Bardia and Mokherjee, Attreyee and Xu, Chongyang and Bindschaedler, Laurent and Tandon, Niket and Arora, Akhil},
  booktitle = {Findings of the Association for Computational Linguistics: EMNLP},
  year = {2025},
}

arxiv

ReasonBench: Benchmarking the (In)Stability of LLM Reasoning

Nearchos Potamitis, Lars Klein, and Akhil Arora

arXiv preprint, 2025

Bib

@article{potamitis2026reasonbench,
  title = {ReasonBench: Benchmarking the (In)Stability of LLM Reasoning},
  author = {Potamitis, Nearchos and Klein, Lars and Arora, Akhil},
  journal = {arXiv preprint},
  year = {2025},
}