What question did this study set out to answer?

This work aims to create a formal framework that addresses trust and auditability in enterprise AI systems by redefining LLMs.

June 30, 20260 citationsOpen Access

View Full Paper

The AI Operating System: Formal Invariants and Execution Guarantees for Bounded Non-Deterministic Computation (ARS-1.0)

ATAnteneh Tessema

Key Points

This work aims to create a formal framework that addresses trust and auditability in enterprise AI systems by redefining LLMs.
Introduced a formal systems engineering framework for AI.
Defined LLMs as probabilistic coprocessors within a deterministic control plane.
Addressed critical enterprise challenges including auditability and scalability.
Demonstrated improved auditability in AI systems, enhancing trust.
Proved that bounded non-deterministic computation could meet enterprise standards.
Validated the framework's effectiveness in controlling probabilistic behavior.

Abstract

A formal systems engineering framework that redefines the LLM as an untrusted probabilistic coprocessor bounded by a strictly deterministic control plane, addressing gaps in enterprise AI such as auditability, scale, and trust

AI से पूछें

Bookmark

View Full Paper

Cite This Study

Anteneh Tessema (Sun,) studied this question.

synapsesocial.com/papers/6a435c38759b888809a52be5 https://doi.org/https://doi.org/10.5281/zenodo.21013544

AI से पूछें

Bookmark

View Full Paper