5 papers
Data Augmentation via Causal-Residual Bootstrapping
Mateusz Gajewski, Sophia Xiao, Bijan Mazaheri
Data augmentation integrates domain knowledge into a dataset by making domain-informed modifications to existing data points. For example, image data can be augmented by duplicatin…
Amortized Causal Discovery with Prior-Fitted Networks
Mateusz Sypniewski, Mateusz Olko, Mateusz Gajewski +1
In recent years, differentiable penalized likelihood methods have gained popularity, optimizing the causal structure by maximizing its likelihood with respect to the data. However,…
VARSHAP: Addressing Global Dependency Problems in Explainable AI with Variance-Based Local Feature Attribution
Mateusz Gajewski, Mikołaj Morzy, Adam Karczmarz +1
Existing feature attribution methods like SHAP often suffer from global dependence, failing to capture true local model behavior. This paper introduces VARSHAP, a novel model-agnos…
Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery
Mateusz Olko, Mateusz Gajewski, Joanna Wojciechowska +3
Neural causal discovery methods have recently improved in terms of scalability and computational efficiency. However, our systematic evaluation highlights significant room for impr…
Accurate estimation of feature importance faithfulness for tree models
Mateusz Gajewski, Adam Karczmarz, Mateusz Rapicki +1
In this paper, we consider a perturbation-based metric of predictive faithfulness of feature rankings (or attributions) that we call PGI squared. When applied to decision tree-base…