6 papers
Taking the Whys Seriously: Limitations of Counterfactual Explanations in Justification and Recourse
Mattia Cerrato, Otto Sahlgren, Xenia Heilmann
Counterfactual explanations (CEs) are widely used in explainable artificial intelligence (AI) to show how a model's outputs would change if the input features were manipulated. Thi…
Are Algorithm Registers Transparent? Perspectives from Germany
Iman Peljto, Xenia Heilmann, Mattia Cerrato
Algorithm registers are public-facing databases that display basic information about algorithms employed in public administration. While several such registers exist across Europe…
From If-Statements to ML Pipelines: Revisiting Bias in Code-Generation
Minh Duc Bui, Xenia Heilmann, Mattia Cerrato +2
Prior work evaluates code generation bias primarily through simple conditional statements, which represent only a narrow slice of real-world programming and reveal solely overt, ex…
Rashomon Sets and Model Multiplicity in Federated Learning
Xenia Heilmann, Luca Corbucci, Mattia Cerrato
The Rashomon set captures the collection of models that achieve near-identical empirical performance yet may differ substantially in their decision boundaries. Understanding the di…
N-Parties Private Structure and Parameter Learning for Sum-Product Networks
Xenia Heilmann, Ernst Althaus, Mattia Cerrato +3
A sum-product network (SPN) is a graphical model that allows several types of probabilistic inference to be performed efficiently. In this paper, we propose a privacy-preserving pr…
FeDa4Fair: Client-Level Federated Datasets for Fairness Evaluation
Xenia Heilmann, Luca Corbucci, Mattia Cerrato +1
Federated Learning (FL) enables collaborative training while preserving privacy, yet it introduces a critical challenge: the "illusion of fairness''. A global model, usually evalua…