Working Paper: Towards Schema-based Learning from a Category-Theoretic Perspective
📰 ArXiv cs.AI
arXiv:2604.10589v1 Announce Type: new Abstract: We introduce a hierarchical categorical framework for Schema-Based Learning (SBL) structured across four interconnected levels. At the schema level, a free multicategory $Sch_{syn}$ encodes fundamental schemas and transformations. An implementation functor $\mathcal{I}$ maps syntactic schemas to representational languages, inducing via the Grothendieck construction the total category $Sch_{impl}$. Implemented schemas are mapped by a functor $Model$
DeepCamp AI