Soft Ontology Priors over Latent Concept Posteriors for Weakly Supervised Neuro-Symbolic Learning
Neuro-symbolic (NeSy) predictors that supervise only the task label suffer from reasoning shortcuts: distinct concept mappings achieve the same maximum likelihood, leaving the recovered concepts unidentifiable and unsafe for downstream reuse.
By Tengwei Song, Robert HoehndorfarXiv - cs.AI
Read the full story
Originally published by arXiv - cs.AI on Oct 23, 2026.
Comments
Sign in to join the discussion.