Testing AI Applications Under Nondeterminism, Drift, and Resource Constraints: A Problem-Driven Multi-Layer Approach

Fuente: Zenodo
Salvato in:
Dettagli Bibliografici
Autore principale: Anonymous
Natura: Recurso digital
Pubblicazione: Zenodo 2025
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866901625619087360
author Anonymous
author_facet Anonymous
contents <p>This repository provides a fully offline, deterministic replication package for the paper <strong>“Testing AI Applications Under Nondeterminism, Drift, and Resource Constraints: A Problem-Driven Multi-Layer Approach.”</strong> It ships lightweight Python implementations of the five-layer AURORA framework (COL, SCL, CDM, RPE, RAS), synthetic datasets, deterministic logs, and scripts that regenerate all reported tables. The package mirrors the AURORA layers and reproduces every reported table from synthetic traces; it does not embed any production integrations or industrial infrastructure.</p> <p> </p> <ul> <li><strong>aurora/</strong> – Core AURORA implementations and utilities.</li> <li><strong>aut_sim/</strong> – Toy industrial simulation components and drift schedules used in the experiments.</li> <li><strong>data/</strong> – Synthetic suites, generated logs, canary scores, paper targets, and small image assets.</li> <li><strong>paper_outputs/tables/</strong> – Pre-generated LaTeX tables (IV–IX) that match the paper targets.</li> <li><strong>scripts/</strong> – Reproduction utilities, deterministic data generators, CI entrypoints, and sanity checks.</li> <li><strong>tests/</strong> – Unit and integration tests that exercise metrics, scheduling, CI artifacts, and reproduction scripts.</li> </ul>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_18001923
institution Zenodo
language
publishDate 2025
publisher Zenodo
record_format zenodo
spellingShingle Testing AI Applications Under Nondeterminism, Drift, and Resource Constraints: A Problem-Driven Multi-Layer Approach
Anonymous
<p>This repository provides a fully offline, deterministic replication package for the paper <strong>“Testing AI Applications Under Nondeterminism, Drift, and Resource Constraints: A Problem-Driven Multi-Layer Approach.”</strong> It ships lightweight Python implementations of the five-layer AURORA framework (COL, SCL, CDM, RPE, RAS), synthetic datasets, deterministic logs, and scripts that regenerate all reported tables. The package mirrors the AURORA layers and reproduces every reported table from synthetic traces; it does not embed any production integrations or industrial infrastructure.</p> <p> </p> <ul> <li><strong>aurora/</strong> – Core AURORA implementations and utilities.</li> <li><strong>aut_sim/</strong> – Toy industrial simulation components and drift schedules used in the experiments.</li> <li><strong>data/</strong> – Synthetic suites, generated logs, canary scores, paper targets, and small image assets.</li> <li><strong>paper_outputs/tables/</strong> – Pre-generated LaTeX tables (IV–IX) that match the paper targets.</li> <li><strong>scripts/</strong> – Reproduction utilities, deterministic data generators, CI entrypoints, and sanity checks.</li> <li><strong>tests/</strong> – Unit and integration tests that exercise metrics, scheduling, CI artifacts, and reproduction scripts.</li> </ul>
title Testing AI Applications Under Nondeterminism, Drift, and Resource Constraints: A Problem-Driven Multi-Layer Approach
url https://doi.org/10.5281/zenodo.18001923