Patryk Marszalek

According to our database, Patryk Marszalek authored at least 11 papers between 2025 and 2026.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of four.

2026

Conceptualizing Embeddings: Sparse Disentanglement for Vision-Language Models.

Piotr Kubaty, Patryk Marszalek, Lukasz Struski, Adam Wróbel, Jacek Tabor, Marek Smieja

CoRR, May, 2026

Bayesian Fine-tuning in Projected Subspaces.

Viktar Dubovik, Patryk Marszalek, Jacek Tabor, Tomasz Kusmierczyk

CoRR, May, 2026

TACTIC for Navigating the Unknown: Tabular Anomaly deteCTion via In-Context inference.

Patryk Marszalek, Tomasz Kusmierczyk, Marek Smieja

CoRR, March, 2026

CounterFlowNet: From Minimal Changes to Meaningful Counterfactual Explanations.

Oleksii Furman, Patryk Marszalek, Jan Maslowski, Piotr Gainski, Maciej Zieba, Marek Smieja

CoRR, February, 2026

HyConEx: Hypernetwork classifier with counterfactual explanations for tabular data.

Patryk Marszalek, Kamil Ksiazek, Oleksii Furman, Ulvi Movsum-zada, Przemyslaw Spurek, Marek Smieja

Neurocomputing, 2026

2025

DiCoFlex: Model-agnostic diverse counterfactuals with flexible control.

Oleksii Furman, Ulvi Movsum-zada, Patryk Marszalek, Maciej Zieba, Marek Smieja

CoRR, May, 2025

HyConEx: Hypernetwork classifier with counterfactual explanations.

Patryk Marszalek, Ulvi Movsum-zada, Oleksii Furman, Kamil Ksiazek, Przemyslaw Spurek, Marek Smieja

CoRR, March, 2025

A Hypernetwork-Based Approach to KAN Representation of Audio Signals.

Patryk Marszalek, Maciej Rut, Piotr Kawa, Piotr Syga

CoRR, March, 2025

ZEUS: Zero-shot Embeddings for Unsupervised Separation of Tabular Data.

Patryk Marszalek, Tomasz Kusmierczyk, Witold Wydmanski, Jacek Tabor, Marek Smieja

Proceedings of the Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2025, 2025

Minimal Ranks, Maximum Confidence: Parameter-efficient Uncertainty Quantification for LoRA.

Patryk Marszalek, Klaudia Balazy, Jacek Tabor, Tomasz Kusmierczyk

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

As Good as It KAN Get: High-Fidelity Audio Representation.

Patryk Marszalek, Maciej Rut, Piotr Kawa, Przemyslaw Spurek, Piotr Syga

Proceedings of the 34th ACM International Conference on Information and Knowledge Management, 2025