Michal Stefánik

53 papers A* 2C 2Misc 1Journal 27Unranked 21

Year	Rank	Type	Title / Venue / Authors
2025	J	`jnl`	Attend or Perish: Benchmarking Attention in Algorithmic Reasoning. CoRR Michal Spiegel, Michal Stefánik, Marek Kadlcík, Josef Kuchar
2025	J	`jnl`	BenCzechMark : A Czech-Centric Multitask and Multimetric Benchmark for Large Language Models with Duel Scoring Mechanism. Trans. Assoc. Comput. Linguistics Martin Fajcik, Martin Docekal, Jan Dolezal, Karel Ondrej, Karel Benes, Jan Kapsa, Pavel Smrz, Alexander Polok, Michal Hradis, Zuzana Neverilová, Ales Horák, Radoslav Sabol, Michal Stefánik, Adam Jirkovsky, David Adamczyk, Petr Hyner, Jan Hula, Hynek Kydlícek
2025	—	`conf`	Can Out-of-Distribution Evaluations Uncover Reliance on Prediction Shortcuts? A Case Study in Question Answering. EMNLP (Findings) Michal Stefánik, Timothee Mickus, Michal Spiegel, Marek Kadlcík, Josef Kuchar
2025	J	`jnl`	Can Out-of-Distribution Evaluations Uncover Reliance on Shortcuts? A Case Study in Question Answering. CoRR Michal Stefánik, Timothee Mickus, Marek Kadlcík, Michal Spiegel, Josef Kuchar
2025	J	`jnl`	Negation: A Pink Elephant in the Large Language Models' Room? CoRR Tereza Vrabcová, Marek Kadlcík, Petr Sojka, Michal Stefánik, Michal Spiegel
2025	A*	`conf`	Pre-trained Language Models Learn Remarkably Accurate Representations of Numbers. EMNLP Marek Kadlcík, Michal Stefánik, Timothee Mickus, Josef Kuchar, Michal Spiegel
2025	J	`jnl`	Pre-trained Language Models Learn Remarkably Accurate Representations of Numbers. CoRR Marek Kadlcík, Michal Stefánik, Timothee Mickus, Michal Spiegel, Josef Kuchar
2025	—	`conf`	Towards the Roots of the Negation Problem: A Multilingual NLI Dataset and Model Scaling Analysis. EMNLP (Findings) Tereza Vrabcová, Marek Kadlcík, Petr Sojka, Michal Stefánik, Michal Spiegel
2025	J	`jnl`	Unravelling the Mechanisms of Manipulating Numbers in Language Models. CoRR Michal Stefánik, Timothee Mickus, Marek Kadlcík, Bertram Højer, Michal Spiegel, Raúl Vázquez, Aman Sinha, Josef Kuchar, Philipp Mondorf
2025	J	`jnl`	VectorEdits: A Dataset and Benchmark for Instruction-Based Editing of Vector Graphics. CoRR Josef Kuchar, Marek Kadlcík, Michal Spiegel, Michal Stefánik
2024	—	`conf`	A Comparative Study of Text Retrieval Models on DaReCzech. RASLAN Jakub Stetina, Martin Fajcik, Michal Stefánik, Michal Hradis
2024	J	`jnl`	A Comparative Study of Text Retrieval Models on DaReCzech. CoRR Jakub Stetina, Martin Fajcik, Michal Stefánik, Michal Hradis
2024	J	`jnl`	BenCzechMark : A Czech-centric Multitask and Multimetric Benchmark for Large Language Models with Duel Scoring Mechanism. CoRR Martin Fajcik, Martin Docekal, Jan Dolezal, Karel Ondrej, Karel Benes, Jan Kapsa, Pavel Smrz, Alexander Polok, Michal Hradis, Zuzana Neverilová, Ales Horák, Radoslav Sabol, Michal Stefánik, Adam Jirkovsky, David Adamczyk, Petr Hyner, Jan Hula, Hynek Kydlícek
2024	—	`conf`	Concept-aware Data Construction Improves In-context Learning of Language Models. ACL (Findings) Michal Stefánik, Marek Kadlcík, Petr Sojka
2024	J	`jnl`	Concept-aware Data Construction Improves In-context Learning of Language Models. CoRR Michal Stefánik, Marek Kadlcík, Petr Sojka
2024	—	`conf`	Self-training Language Models for Arithmetic Reasoning. EMNLP (Findings) Marek Kadlcík, Michal Stefánik
2024	J	`jnl`	Self-training Language Models for Arithmetic Reasoning. CoRR Marek Kadlcík, Michal Stefánik
2024	—	`conf`	Think Twice: Measuring the Efficiency of Eliminating Prediction Shortcuts of Question Answering Models. EACL (1) Lukás Mikula, Michal Stefánik, Marek Petrovic, Petr Sojka
2023	J	`jnl`	Applications of deep language models for reflective writings. Educ. Inf. Technol. Jan Nehyba, Michal Stefánik
2023	—	`conf`	CIVQA: Czech Invoice Visual Question Answering. RASLAN Sárka Scavnická, Michal Stefánik, Petr Sojka
2023	A*	`conf`	Calc-X and Calcformers: Empowering Arithmetical Chain-of-Thought through Interaction with Symbolic Systems. EMNLP Marek Kadlcík, Michal Stefánik, Ondrej Sotolár, Vlastimil Martinek
2023	J	`jnl`	Calc-X: Enriching Arithmetical Chain-of-Thoughts Datasets by Interaction with Symbolic Systems. CoRR Marek Kadlcík, Michal Stefánik
2023	J	`jnl`	Concept-aware Training Improves In-context Learning Ability of Language Models. CoRR Michal Stefánik, Marek Kadlcík
2023	—	`conf`	People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts. ACL (Findings) Vit Novotny, Kristina Luger, Michal Stefánik, Tereza Vrabcová, Ales Horák
2023	J	`jnl`	People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts. CoRR Vít Novotný, Kristýna Luger, Michal Stefánik, Tereza Vrabcová, Ales Horák
2023	J	`jnl`	Resources and Few-shot Learners for In-context Learning in Slavic Languages. CoRR Michal Stefánik, Marek Kadlcík, Piotr Gramacki, Petr Sojka
2023	—	`conf`	Soft Alignment Objectives for Robust Adaptation of Language Generation. ACL (1) Michal Stefánik, Marek Kadlcík, Petr Sojka
2023	J	`jnl`	Think Twice: Measuring the Efficiency of Eliminating Prediction Shortcuts of Question Answering Models. CoRR Lukás Mikula, Michal Stefánik, Marek Petrovic, Petr Sojka
2022	J	`jnl`	AdaptOr: Objective-Centric Adaptation Framework for Language Models. CoRR Michal Stefánik, Vít Novotný, Nikola Groverová, Petr Sojka
2022	—	`conf`	Adaptor: Objective-Centric Adaptation Framework for Language Models. ACL (demo) Michal Stefánik, Vít Novotný, Nikola Groverová, Petr Sojka
2022	—	`conf`	Combining Sparse and Dense Information Retrieval. CLEF (Working Notes) Vít Novotný, Michal Stefánik
2022	—	`conf`	Diverse Semantics Representation is King. CLEF (Working Notes) Martin Geletka, Vojtech Kalivoda, Michal Stefánik, Marek Toma, Petr Sojka
2022	—	`conf`	Information Extraction from Business Documents: A Case Study. RASLAN Martin Geletka, Mikulás Bankovic, Dávid Melus, Sárka Scavnická, Michal Stefánik, Petr Sojka
2022	—	`conf`	Methods for Estimating and Improving Robustness of Language Models. NAACL-HLT (Student Research Workshop) Michal Stefánik
2022	J	`jnl`	Methods for Estimating and Improving Robustness of Language Models. CoRR Michal Stefánik
2022	J	`jnl`	Soft Alignment Objectives for Robust Adaptation in Machine Translation. CoRR Michal Stefánik, Marek Kadlcík, Petr Sojka
2022	—	`conf`	Towards General Document Understanding through Question Answering. RASLAN Sárka Scavnická, Michal Stefánik, Marek Kadlcík, Martin Geletka, Petr Sojka
2022	J	`jnl`	What is Not in the Context? Evaluation of Few-shot Learners with Informative Demonstrations. CoRR Michal Stefánik, Marek Kadlcík
2022	J	`jnl`	When FastText Pays Attention: Efficient Estimation of Word Representations using Constrained Positional Weighting. J. Univers. Comput. Sci. Vít Novotný, Michal Stefánik, Eniafe Festus Ayetiran, Petr Sojka, Radim Rehurek
2021	C	`conf`	CICM'21 Systems Entries. CICM Martin Líska, Dávid Lupták, Vít Novotný, Michal Ruzicka, Boris Shminke, Petr Sojka, Michal Stefánik, Makarius Wenzel
2021	—	`conf`	Ensembling Math Information Retrieval Systems: MIRMU and MSM at ARQMath 2021. CLEF (Working Notes) Vít Novotný, Michal Stefánik, Dávid Lupták, Martin Geletka, Petr Zelina, Petr Sojka
2021	J	`jnl`	One Size Does Not Fit All: Finding the Optimal N-gram Sizes for FastText Models across Languages. CoRR Vít Novotný, Eniafe Festus Ayetiran, Dávid Lupták, Michal Stefánik, Petr Sojka
2021	Misc	`conf`	One Size Does Not Fit All: Finding the Optimal Subword Sizes for FastText Models across Languages. RANLP Vít Novotný, Eniafe Festus Ayetiran, Dalibor Bacovský, Dávid Lupták, Michal Stefánik, Petr Sojka
2021	—	`conf`	Regressive Ensemble for Machine Translation Quality Evaluation. WMT@EMNLP Michal Stefánik, Vít Novotný, Petr Sojka
2021	J	`jnl`	Regressive Ensemble for Machine Translation Quality Evaluation. CoRR Michal Stefánik, Vít Novotný, Petr Sojka
2021	—	`conf`	Towards Domain Robustness of Neural Language Models. RASLAN Michal Stefánik, Petr Sojka
2021	C	`conf`	WebMIaS on Docker - Deploying Math-Aware Search in a Single Line of Code. CICM Dávid Lupták, Vít Novotný, Michal Stefánik, Petr Sojka
2021	J	`jnl`	WebMIaS on Docker: Deploying Math-Aware Search in a Single Line of Code. CoRR Dávid Lupták, Vít Novotný, Michal Stefánik, Petr Sojka
2021	J	`jnl`	When FastText Pays Attention: Efficient Estimation of Word Representations using Constrained Positional Weighting. CoRR Vít Novotný, Michal Stefánik, Eniafe Festus Ayetiran, Petr Sojka
2020	J	`jnl`	Text classification with word embedding regularization and soft similarity measure. CoRR Vít Novotný, Eniafe Festus Ayetiran, Michal Stefánik, Petr Sojka
2020	—	`conf`	Three is Better than One: Ensembling Math Information Retrieval Systems. CLEF (Working Notes) Vít Novotný, Petr Sojka, Michal Stefánik, Dávid Lupták
2020	—	`conf`	Towards Useful Word Embeddings. RASLAN Vít Novotný, Michal Stefánik, Dávid Lupták, Petr Sojka
2019	—	`conf`	Quo Vadis, Math Information Retrieval. RASLAN Petr Sojka, Vít Novotný, Eniafe Festus Ayetiran, Dávid Lupták, Michal Stefánik