Author name cluster

Tomasz Paweł Michalak

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.

2 papers

1 author row

AAAI Conference 2026 Conference Paper

On Stealing Graph Neural Network Models

Marcin Podhajski
Jan Dubiński
Franziska Boenisch
Adam Dziedzic
Agnieszka Pręgowska
Tomasz Paweł Michalak

Current graph neural network (GNN) model-stealing methods rely heavily on queries to the victim model, assuming no hard query limits. However, in reality, the number of allowed queries can be severely limited. In this paper, we demonstrate how an adversary can extract a GNN with very limited interactions with the model. Our approach first enables the adversary to obtain the model backbone without making direct queries to the victim model and then to strategically utilize a fixed query limit to extract the most informative data. The experiments on eight real-world datasets demonstrate the effectiveness of the attack, even under a very restricted query limit and under defense against model extraction in place. Our findings underscore the need for robust defenses against GNN model extraction threats.

PDF Details DOI

IJCAI Conference 2015 Conference Paper

The Game-Theoretic Interaction Index on Social Networks with Applications to Link Prediction and Community Detection

Piotr Lech Szczepański
Aleksy Stanisław Barcz
Tomasz Paweł Michalak
Talal Rahwan

Measuring similarity between nodes has been an issue of extensive research in the social network analysis literature. In this paper, we construct a new measure of similarity between nodes based on the game-theoretic interaction index (Grabisch and Roubens, 1997). Despite the fact that, in general, this index is computationally challenging, we show that in our network application it can be computed in polynomial time. We test our measure on two important problems, namely link prediction and community detection, given several real-life networks. We show that, for the majority of those networks, our measure outperforms other local similarity measures from the literature.

PDF Details