Profile I am interested in building AI systems that can plan, adapt, and generalise in complex environments. My research focuses on AI planning, reinforcement learning, abstraction, and decision-making under distribution shift.
I hold an M.Sc. in Machine Learning (2023), supervised by Prof. Piotr Miłoś and Prof. Marek Cygan, and a B.Sc. in Computer Science (2021), supervised by Prof. Henryk Michalewski and Prof. Łukasz Kaiser, both from the University of Warsaw, MIM Faculty. I currently work as a Senior AI Consultant at EY on agentic AI systems.
I hold an M.Sc. in Machine Learning (2023), supervised by Prof. Piotr Miłoś and Prof. Marek Cygan, and a B.Sc. in Computer Science (2021), supervised by Prof. Henryk Michalewski and Prof. Łukasz Kaiser, both from the University of Warsaw, MIM Faculty. I currently work as a Senior AI Consultant at EY on agentic AI systems.
Alongside this, I have gained experience in research and engineering roles at NVIDIA, Microsoft, Samsung, DeepFlare, and EY. Since 2020, I have also been involved in the ML in PL Association, supporting scientific programme activities and conference organisation across several editions, including its upcoming 10th anniversary conference.
You can view my full CV here.
Selected Publications
Resolution of Recursive Data Corruption to Transform T-cell Epitope Discovery
bioRxiv 2026
bioRxiv 2026
Paper
Cite
Preibisch, G., Tyrolski, M., Kucharski, P., Giziński, S., Grzegorczyk, P., Moon, S., Kim, S., Zaro, B., & Gambin, A. (2026). Resolution of recursive data corruption to transform T-cell epitope discovery. bioRxiv, 2026.03.30.710191. https://doi.org/10.64898/2026.03.30.710191
bioRxiv preprint on data contamination in T-cell epitope discovery; reframes MHC-I prediction as protein-centric learning to rank.
OpenGVL: Benchmarking Visual Temporal Progress for Data Curation
CoRL 2025 workshop
CoRL 2025 workshop
arXiv GitHub Repo Live Benchmark
Cite
Budzianowski, P., Wiśnios, E., Tyrolski, M., Góral, G., Kulakov, I., Petrenko, V., & Walas, K. (2025). OpenGVL--Benchmarking Visual Temporal Progress for Data Curation. arXiv preprint arXiv:2509.17321.
Benchmark, toolkit, and live leaderboard for evaluating VLM temporal progress estimation in robotics videos; supports automated dataset curation.
What Matters in Hierarchical Search for Combinatorial Reasoning Problems?
ICLR 2024 (Generative Models for Decision Making)
ICLR 2024 (Generative Models for Decision Making)
Paper Code
Cite
Zawalski, M., Góral, G., Tyrolski, M., Wiśnios, E., Budrowski, F., Cygan, M., Kuciński, Ł., & Miłoś, P. (2024). What matters in hierarchical search for combinatorial reasoning problems? arXiv preprint arXiv:2406.03361.
Empirical study of when hierarchical search helps in combinatorial reasoning, with evaluation guidelines for comparing methods.
Adjusting Planning Horizon with Adaptive Subgoal Search
ICLR 2023 (Top-5%, Oral)
ICLR 2023 (Top-5%, Oral)
Paper Code Website Colab
Cite
Zawalski, M., Tyrolski, M., Czechowski, K., Odrzygóźdź, T., Stachura, D., Piękos, P., Wu, Y., Kuciński, Ł., & Miłoś, P. (2022). Fast and precise: Adjusting planning horizon with adaptive subgoal search. arXiv preprint arXiv:2206.00702.
AdaSubS adjusts planning horizon using generated subgoals and reachability filtering; evaluated on Sokoban, Rubik's Cube, and INT.
Hierarchical Transformers Are More Efficient Language Models
NAACL 2022
NAACL 2022
Paper
Cite
Nawrot, P., Tworkowski, S., Tyrolski, M., Kaiser, Ł., Wu, Y., Szegedy, C., & Michalewski, H. (2022). Hierarchical transformers are more efficient language models. Findings of the Association for Computational Linguistics: NAACL 2022, 1559-1571.
Hourglass adds downsampling and upsampling to Transformers for more efficient long-sequence modeling; evaluated on ImageNet32 and enwik8.
Other Papers
Explainable Machine Learning at Microsoft MLADS
Microsoft MLADS 2022
Microsoft MLADS 2022
Cite
Tyrolski, M. (2022). Explainable machine learning at Microsoft MLADS (Internal Microsoft MLADS Conference presentation).
Internal Microsoft MLADS presentation on explainable machine learning work during a Microsoft Ireland internship.
Enhancing Antigenic Peptide Discovery
ICLR 2023, MLDD Workshop
ICLR 2023, MLDD Workshop
Paper
Cite
Giziński, S., Preibisch, G., Kucharski, P., Tyrolski, M., Rembalski, M., Grzegorczyk, P., & Gambin, A. (2024). Enhancing antigenic peptide discovery: Improved MHC-I binding prediction and methodology. Methods, 224, 1–9. https://doi.org/10.1016/j.ymeth.2024.01.016
Study of evaluation pitfalls in MHC-I presentation prediction, with a unified methodology and model for peptide-MHC-I binding.
Selected Projects
CaRL Library: Combinatorial RL for Planning
CaRL is an open-source library for offline and online reinforcement/imitation learning in combinatorial planning.
Todoist Assistant
Local-first Todoist dashboard and automation toolkit with optional summaries and read-only chat over cached activity.
