Technical Record: Individualized Semantic Reference Layers for Language Models — Dated Description and Anchor Chain (September 2026)

Dated technical record describing a method in which a language model's own definitions serve as a reference layer, against which one person's meanings and decision reasons are elicited, recorded in versioned form and made usable by later AI systems. The record contains a frozen technical description, a research record and an anchor chain linking earlier dated work by the author. Files are access-restricted; integrity can be verified through the SHA-256 values below. Evidence Package (ZIP) SHA-256: a9290946dfc28fdf8088d2e9b76356b8302628f9ab1adb06f6f9804425ff51fcCompanion Anchor Chain (ZIP) SHA-256: 3f31eaa16a394647cb07c9f09070c04e51ec2161faa01517bf6d91152f1e0bee