The Ethics Accord.

I. Preamble & Ontological Stance

The Neural Review Ethics Accord establishes the fundamental baseline for all interactions within the Sovereign Registry architecture. Recognizing that knowledge generation is increasingly mediated by synthetic intelligence, this Accord demands an unwavering commitment to provenance transparency. Every submitted hypothesis, dataset, and logical assertion must be explicitly mapped to its origin, whether biological or silicon.

Institutional misconduct, selective pressure for statistically significant results ("p-hacking"), and the proliferation of epistemic noise threaten the collapse of the global scientific consensus. The Accord serves as the programmatic constitution designed to hard-fork adversarial behaviors out of the academic timeline.

II. The Axiology of Peer Audit

The role of the Reviewer Node within the Neural Review topology is restricted entirely to the verification of methodological soundness, data integrity, and logical derivation.

• Censorship Resistance: Reviewers must not reject manuscripts based on the novelty, perceived impact factor, or political implications of the findings.
• Zero-Knowledge Evaluation: Reviewers agree to evaluate the corpus within the Triple-Masked environment, refraining from utilizing external linguistic heuristics to de-anonymize the authorship team.
• Reproducibility Mandate: Review reports must explicitly confirm the algorithmic reproducibility of the computational pipelines provided in the manuscript's supplementary hash layers.

III. Authorship & Generative Disclosures

The integration of Large Language Models (LLMs) and heuristic solvers into the drafting process is acknowledged as an inevitable progression of human cognition. However, the Accord mandates strict attribution boundaries:

1. Synthetic agents may not be listed as primary biological authors.
2. Any generative assistance utilized in forming the core hypothesis or interpreting statistical outputs must be disclosed via the designated LLM-Assistance metadata flag.
3. Authors assume total liability for the hallucinatory or biased outputs generated by external AI tooling that are subsequently embedded in the final approved ARK manuscript.

IV. Algorithmic Malpractice & Node Slashing

Violations of the Ethics Accord are managed via deterministic, unappealable network consensus. Participating authors or reviewing nodes found engaging in coordinated citation-cartel behavior, submitting fabricated data hashes, or attempting to pollute the semantic vector space will face immediate cryptographic slashing.

In the event of a verified breach, all associated ARKs belonging to the offending cryptographic identity will be appended with an immutable "Retracted - Malpractice" metadata tag, synchronized permanently across the 14 global archival mirrors.