arXiv:2604.04344v3 Announce Type: replace
Abstract: We establish a computation-substrate-agnostic inference architecture in which domain is an explicit first-class computational parameter. This produces domain-scoped pruning that reduces per-query sea…
arXiv:2604.04344v2 Announce Type: replace
Abstract: We establish a computation-substrate-agnostic inference architecture in which domain is an explicit first-class computational parameter. This produces domain-scoped pruning that reduces per-query sea…
arXiv:2604.01770v2 Announce Type: replace
Abstract: Knowledge graphs store large numbers of relations efficiently, but they remain weak at representing a quieter difficulty: the meaning of a concept often shifts with the domain in which it is used. A …