Author name cluster

Marcin Kurdziel

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

3 papers

1 author row

AAMAS Conference 2025 Conference Paper

Learning Real-Life Approval Elections

Piotr Faliszewski
Lukasz Janeczko
Andrzej Kaczmarczyk
Marcin Kurdziel
Grzegorz Pierczynski
Stanislaw Szufa

We study the independent approval model (IAM) for approval elections, where each candidate has its own approval probability and is approved independently of the other ones. This model generalizes, e. g. , the impartial culture, the Hamming noise model, and the resampling model. We propose algorithms for learning IAMs and their mixtures from data, using either maximum likelihood estimation or Bayesian learning. We then apply these algorithms to a large set of elections from the Pabulib database. In particular, we find that single-component models are rarely sufficient to capture the complexity of real-life data, whereas their mixtures perform well.

PDF

AAAI Conference 2023 Conference Paper

Neural Representations Reveal Distinct Modes of Class Fitting in Residual Convolutional Networks

Michał Jamroż
Marcin Kurdziel

We leverage probabilistic models of neural representations to investigate how residual networks fit classes. To this end, we estimate class-conditional density models for representations learned by deep ResNets. We then use these models to characterize distributions of representations across learned classes. Surprisingly, we find that classes in the investigated models are not fitted in a uniform way. On the contrary: we uncover two groups of classes that are fitted with markedly different distributions of representations. These distinct modes of class-fitting are evident only in the deeper layers of the investigated models, indicating that they are not related to low-level image features. We show that the uncovered structure in neural representations correlate with memorization of training examples and adversarial robustness. Finally, we compare class-conditional distributions of neural representations between memorized and typical examples. This allows us to uncover where in the network structure class labels arise for memorized and standard inputs.

PDF Details DOI

NeurIPS Conference 2020 Conference Paper

A Bayesian Nonparametrics View into Deep Representations

Michał Jamroż
Marcin Kurdziel
Mateusz Opala

We investigate neural network representations from a probabilistic perspective. Specifically, we leverage Bayesian nonparametrics to construct models of neural activations in Convolutional Neural Networks (CNNs) and latent representations in Variational Autoencoders (VAEs). This allows us to formulate a tractable complexity measure for distributions of neural activations and to explore global structure of latent spaces learned by VAEs. We use this machinery to uncover how memorization and two common forms of regularization, i. e. dropout and input augmentation, influence representational complexity in CNNs. We demonstrate that networks that can exploit patterns in data learn vastly less complex representations than networks forced to memorize. We also show marked differences between effects of input augmentation and dropout, with the latter strongly depending on network width. Next, we investigate latent representations learned by standard $\beta$-VAEs and Maximum Mean Discrepancy (MMD) $\beta$-VAEs. We show that aggregated posterior in standard VAEs quickly collapses to the diagonal prior when regularization strength increases. MMD-VAEs, on the other hand, learn more complex posterior distributions, even with strong regularization. While this gives a richer sample space, MMD-VAEs do not exhibit independence of latent dimensions. Finally, we leverage our probabilistic models as an effective sampling strategy for latent codes, improving quality of samples in VAEs with rich posteriors.

PDF Details