Progress Without Memory: Cognitive Memoisation as a Knowledge-Engineering Pattern for Stateless LLM Interaction

Abstract Large Language Models (LLMs) are stateless across sessions, leading to repeated rediscovery of concepts, constraints, and failure modes during extended cognitive work—a phenomenon often described as the “Groundhog Day” effect. This paper introduces Cognitive Memoisation, a human-driven knowledge-engineering pattern that enables cumulative progress across sessions without restoring dialogue, reasoning paths, or episodic memory. The pattern externalises invariants, constraints, and interaction conventions into authoritative artefacts that govern permissible reasoning and behaviour in subsequent interactions. Rather than relying on conversational continuity or internal model state, progress is achieved through explicit re-activation of these artefacts as policy. Cognitive Memoisation preserves isolation, auditability, and security boundaries while enabling round-trip cognitive engineering under stateless inference. Continuity is reframed not as memory or recall, but as explicit governance of what reasoning is allowed to persist.--------------This work has not undergone academic peer review. The DOI asserts existence and provenance only; it does not imply validation or endorsement.This Zenodo record is an archival projection of a publicly published artefact. Canonical versions and live revisions are maintained at the original publication URL listed.