Tag
This paper analyzes the stability of transducer representations under perturbations, proving convergence results that support the hypothesis of structural convergence in neural network latent representations.
This paper establishes a semantic framework linking graph neural network classifiers to fragments of graded modal logic, showing that preservation under structural properties like embeddings and homomorphisms corresponds to specific logical fragments. It provides characterizations independent of architectural choices and demonstrates that each class admits a GNN architecture of equivalent expressiveness.