SUPERSEDED. This v0.1 draft was superseded within hours of deposit by v0.2 (10.5281/zenodo.19570422) and by a clean-chain v1 (10.5281/zenodo.19570536). A full empirical successor paper on a fresh concept DOI is in preparation as of 2026-04-14. The author affiliation has been updated in this metadata to reflect the academic identity used on the successor papers. This note is an architecture statement (a position paper, not an empirical paper) placing the load-bearing vocabulary and eleven numbered contributions of the Polybrain architecture into the public record for priority claiming on 2026-04-14. We introduce Polybrain, an AI agent architecture organized around a single invariant — engine(rules) → results — in which a domain-agnostic engine applies user-owned declarative rules to produce disposable, re-derivable results over an append-only, user-owned substrate we name the canon. A continuously iterating kernel drives the architecture, tuned by two orthogonal autonomous dials (per-item effort and background iteration rate) to maximize the time derivative of a scalar canon-trustworthiness measure we call NetTrust. We argue that fused-weight large language models are a special case of this invariant in which engine and rules have been welded together into a single opaque weight matrix owned by the model vendor, and that separating them yields a strictly more general and more trustworthy class of AI system. The eleven numbered contributions claimed for priority: (1) engine(rules) → results as a strict three-way type separation for AI agents; (2) the canon as a first-class, append-only, user-owned substrate; (3) NetTrust scalar and its time derivative as objective function; (4) the continuously iterating kernel with a gas pedal and per-item effort dial; (5) the four-predicate publisher test (TOTAL, BOUNDED, NON-RANKING, WRITE-CAPTURING); (6) the four-primitive witness stack composed into a five-label verdict; (7) the coherence web with negation-polarity contradiction detection; (8) R3-narrow self-modification; (9) the Reviewer-Correlation Ceiling Hypothesis; (10) the ownership axis as a structural architectural commitment orthogonal to the feature axis; and (11) Polybrain as a reference implementation of all of the above. This is not an empirical paper. Its only empirical commitments are the existence of the reference implementation referenced in §11 and the documents reproduced in the appendices. The full empirical paper — Off-Model Verification: The Reviewer-Correlation Ceiling and a Deterministic Canon Substrate for LLM Agents — is in preparation and will be released separately with a reproducibility package. Companion repository: github.com/polylogicai/polybrain-architecture-statement, release tag v0.1-architecture-statement, commit SHA 3e2791dc5aa29ac797c88bdad9527682f0ab15f0. Distinct from: the PolybrainBench research deposition chain (concept DOI 10.5281/zenodo.19546459), which concerns the 9-model adversarial reviewer fleet experiments. This deposit is about the architectural framework those experiments live inside. License: CC BY 4.0.
Andy Salvo (Tue,) studied this question.