Perception
Reference: This page is a source-based technical synthesis of the LaTeX chapter cited below. For exact syntax, availability or ABI signatures, verify the versioned source, manifest and executable tests.
Scope and source boundary
Agent perception may yield observations or candidate relations; the source does not grant observations the status of proven facts automatically.
Conceptual memory provides structured context and provenance for knowledge-driven applications, including agents. It is not interchangeable with a conversation transcript, cached embedding, session buffer or invented “remembered” fact. The model distinguishes stable semantic relations from derived or temporary projections.
Engineering rules
- Give facts an explicit conceptual identity and scope.
- Record where observations come from and how their validity is established.
- Refresh derived views and indexes when authoritative relations change.
- Keep session memory separate from durable model and semantic memory.
Chapter outline (original LaTeX headings)
- Purpose
- Beyond Document Retrieval
- Structured Context for AI
- Actions as Safe Execution Handles
- Conceptual Memory and Explanations
- Design Rules
LaTeX provenance
Primary chapter: Declarative Conceptual Programming with the logiCells Platform/developer-guide/chapters/conceptual-memory-for-ai-applications.tex.