About Me

I am a postdoctoral researcher at the Institute of Formal and Applied Linguistics, Faculty of Mathematics and Physics, Charles University and a faculty member at the Division of Machine Learning, Institute of Computing Science, Poznan University of Technology. I received my Ph.D. in Computer Science from Poznań University of Technology under the supervision of prof. Jerzy Stefanowski in 2022. Currently, I am doing my postdoc with prof. Ondřej Dušek.

My research interests include:

  • explainable AI/ML/NLP
  • (accurate, controllable) natural language generation
  • explainable sentiment analysis
  • learning from imbalanced and complex data (in the past)

I was an investigator in grants of the European Research Council, Polish National Science Center, National Centre for Research and Development, Norway Funds, EU's HORIZON 2020 grants as well as in R&D research projects for industry. I made research stays at Charles University in Prague and Ljubljana University. From 2019 to 2023, I was a machine learning consultant at the Poznan Supercomputing and Networking Centre (Polish Academy of Sciences). Before joining academia, I worked in Human Language Technologies Lab of Samsung R&D, in Data Science team of Berlin-based startup Roq.ad and in Comarch.

I was an Area Chair for ACL ARR (ACL, EMNLP, EACL, ..), NeurIPS and GEM workshop. I was in the Programme Committees of ACL, EMNLP, NeurIPS, ICML, ICLR, EACL, NAACL, ECML PKDD, ECAI, COLING, INLG, SIGDIAL, LREC and multiple workshops (HumEval, YNLG, GEM, ...). I served as a reviewer for Machine Learning (Springer), Pattern Recognition (Springer), Neurocomputing (Elsevier), IEEE Transactions on Knowledge and Data Engineering, Expert Systems With Applications (Elsevier), Mateheuristics (Springer), Journal of Experimental and Theoretical Artificial Intelligence (Taylor&Francis), DSH: Digital Scholarship in the Humanities, AAAI Conference on Artificial Intelligence, SIAM Data Mining, International Journal of Human-Computer Interaction (Taylor&Francis), Informatics in Medicine Unlocked (Elsevier), International Journal of Applied Mathematics and Computer Science (De Gruyter), Foundations of Computing and Decision Sciences (De Gruyter), ICAISC.

I was also a tutor and founder of Group of Horribly Optimistic STatisticians (2017-2022), initiator and main organizer of GHOST Day: Applied Machine Learning Conference (2019-2022), co-organizer of 1st Polish Artificial Intelligence Conference and The 2nd Workshop on Practical LLM-assisted Data-to-Text Generation. I am a member of Association for Computational Linguistics, European Association for Machine Translation and Polish Special Interest Group on Machine Learning.

I was (co-)supervisor of following Master's thesis:

  • Neural architectures for diffusion large language models (tentative), Marcin Kapiszewski
  • Towards better evaluation of LLM agents enviroments (tentative), Jan Orłowski
  • Personalized textual style transfer for Polish news portals (tentative), Jakub Bilski
  • Non-autoregressive text generation with Kuramoto oscillatory networks (tentative), Wiktor Kamzela
  • Disentangling knowledge and linguistic abilities in large language models (tentative), Filip Baranowski
  • Interpretable methods for data-to-text generation (2026), Dominik Maćkowiak
  • Non-autoregressive methods for natural language generation, Jędrzej Warczyński (2025) [see our papers: 1, 2]
  • Multilingual knowledge transfer techniques for large language models, Antoni Lasik (2025)
  • Explainable LLM-based evaluation for NLG using error analysis, Ivan Kartáč (2025) [our paper]
  • Explaining image classification in natural language, Adam Wojciechowski (2023) [our paper][2nd place in "The Best Polish CS Master's thesis" competition]
  • Neural methods for answering questions with incomplete problem specifications, Bogumiła Walkowiak (2023)
  • Natural language processing methods for neutralising clickbaits in social networks, Mateusz Woźny (2023) [see our papers: 1 2] [distinction in "The Best Polish CS Master's thesis" competition] [award of the City of Poznań for the best Master's thesis]
  • Gradient boosting trees for multiclass imbalanced data, Stanisław Kowalczyk (2023)
  • Explaining neural recommendation systems in natural language, Jakub Raczyński (2022) [award of the City of Poznań for an outstanding Master's thesis] [see our paper]
  • Automatic extraction of aspect and opinion phrases in text sentiment analysis, Iwo Naglik (2022) [TOP5 PUT alumnus] [see our papers: 1 2]
  • Federated learning-based classification model, Kamil Basiukajc (2022)
  • Prototypical Convolutional Neural Network for a phrase-based explanation of sentiment classification, Kamil Pluciński (2021) [see our paper] [distiction in National Competition for an Outstanding Thesis in CS/EE]
  • Similarity learning as a feature transform for multi-class imbalanced classification, Damian Horna (2021) [TOP5 PUT alumnus] [see our paper]
  • Unsupervised abstractive text summarization with theoretical model of importance, Mateusz Skiba (2020)
  • Time aspect in making actionable predictions of a conversation breakdown, Piotr Janiszewski (2020) [see our paper]