5 papers
The Boolean Power of ReLU
Pablo Barceló, Floris Geerts, Matthias Lanzinger +2
We prove that, on finite simple undirected graphs equipped with a single Boolean node feature, the Boolean queries expressible in -MPLang, for any collection of eventually c…
A Logical View of GNN-Style Computation and the Role of Activation Functions
Pablo Barceló, Floris Geerts, Matthias Lanzinger +2
We study the numerical and Boolean expressiveness of MPLang, a declarative language that captures the computation of graph neural networks (GNNs) through linear message passing and…
Message Passing on the Edge: Towards Scalable and Expressive GNNs
Pablo Barceló, Fabian Jogl, Alexander Kozachinskiy +3
Graph neural networks (GNNs) are widely used in graph learning and most architectures propagate information by passing messages between vertices. In this work, we shift our attenti…
Language Generation: Complexity Barriers and Implications for Learning
Marcelo Arenas, Pablo Barceló, Luis Cofré +1
Kleinberg and Mullainathan showed that language generation in the limit is always possible at the level of computability: given enough positive examples, a learner can eventually g…
When is the Computation of a Feature Attribution Method Tractable?
P. Barceló, R. Cominetti, M. Morgado
Feature attribution methods have become essential for explaining machine learning models. Many popular approaches, such as SHAP and Banzhaf values, are grounded in power indices fr…