Learning with Logical Constraints but without Shortcut Satisfaction

arXiv:2403.00329v1 Announce Type: cross
Abstract: Recent studies in neuro-symbolic learning have explored the integration of logical knowledge into deep learning via encoding logical constraints as an additional loss function. However, existing approaches tend to vacuously satisfy logical constraints through shortcuts, failing to fully exploit the knowledge. In this paper, we present a new framework for learning with logical constraints. Specifically, we address the shortcut satisfaction issue by introducing dual variables for logical connectives, encoding how the constraint is satisfied. We …

abstract arxiv constraints cs.ai cs.lg deep learning encoding exploit function integration knowledge loss neuro paper shortcut studies through type via