Public Archive
Public Research Documents
A curated selection of sanitized documents. Internal audit logs, training scripts, raw diagnostic outputs, and full pipeline data are not published here. Only documents that stand alone without misleading context are released.
⚠ All documents below are sanitized versions of internal materials. Sensitive implementation details, full audit trails, and private research notes have been redacted.
NMU Architecture: Conceptual Overview
A high-level explanation of the four-layer NINMENI architecture. Covers NMU formation principles, the epistemic motivation for each stage, and why this design diverges from transformer-based prediction paradigms. Written for readers familiar with language model research.
Read full architecture page →200k Shadow Pilot: Summary Report
Summary of the 200k NMU shadow pilot run completed June 12, 2026. Covers: 194,297 NMUs formed, 21,587 records processed, all epistemic audit checks passed (MRO conflict = 0, true contradictions = 0, determinism PASS, source fidelity PASS, renderer inspection PASS). Internal diagnostics and training scripts are redacted.
Read status dashboard →On Scale-Up and Content Calibration at 500k Staging
An honest analysis of what was observed during the 500k staging phase. Tooling damage incident, content calibration, and projection quantity gates required recovery protocol. This note explains what that means architecturally, and what the preparation and re-verification entail. No internal audit data included.
Read timeline entry →Responsible Research Boundaries
A formal statement of what NINMENI claims and does not claim. This document exists to prevent misrepresentation of internal milestones as external benchmark results or capability claims.
- Not AGI or general intelligence.
- Not a production-ready system.
- Not a wrapper or fine-tune of an existing LLM.
- Not a public chatbot or accessible AI product.
- Not claiming superiority over any other AI system on any task.
- Not validated by external review, benchmark, or peer-reviewed paper (yet).
Creator & Research Lineage
The origin context of the NINMENI research project: who is behind it, where the intellectual lineage comes from, and why an independent Indonesia-native project is the right place to ask these questions.
Read lineage page →