Zero State Kernel:Axiomatic Foundations of General Intelligence

Read full text Summarise with VIRAT
Authors: Zhiye Zhang
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This work establishes the first axiomatic closure for General Intelligence, replacing empirical scaling with structural necessity. Current AGI relies on black-box statistical models facing unavoidable crises: hallucination, value drift, cognitive closure, and runaway risks, all due to the lack of an absolute reference frame. We propose the Zero State Kernel, a mathematically consistent, physically grounded foundation centered on a hardware-rooted zero-state reference and structural deviation minimization. Intelligence is defined as reducing internal deviation from this invariant frame. The unified architecture includes six layers and a stable Z‑E‑R‑S‑H loop. A dual‑threshold safety mechanism and five formal theorems eliminate hallucinations and ensure universal safety. This framework transforms AGI from empirical speculation into rigorous, verifiable, and reliably engineerable science.

Fields of Study & Subjects

Psychiatry Mental Health Neuroscience Artificial Intelligence Applications Cognitive Computing and Networks

Publication details

Journal / source
Figshare
Publication date
1 January 2026
Language
EN
Document type
Thesis