On the Independence Assumption in Neurosymbolic Learning

arXiv:2404.08458v1 Announce Type: cross
Abstract: State-of-the-art neurosymbolic learning systems use probabilistic reasoning to guide neural networks towards predictions that conform to logical constraints over symbols. Many such systems assume that the probabilities of the considered symbols are conditionally independent given the input to simplify learning and reasoning. We study and criticise this assumption, highlighting how it can hinder optimisation and prevent uncertainty quantification. We prove that loss functions bias conditionally independent neural networks to become overconfident in their predictions. As …

abstract art arxiv constraints cs.ai cs.lg guide independent learning systems networks neural networks predictions reasoning state stat.ml study systems type