About the publication

Research with a clear account of the evidence.

OpenProse Research publishes technical investigations of decision systems, agents, and their evaluation.

Scope

We are interested in how intelligent systems behave under controlled changes to their inputs, representations, and operating conditions. Our reports are intended for researchers and engineers who want to understand a result closely enough to test it or apply it responsibly.

This is an OpenProse publication. An article published here is not, by that fact alone, independently peer reviewed. Each article should state its authorship, relevant affiliations, and review status.

Editorial standards

We use Standard Technical English and make the main finding accessible without weakening the technical account. A report should distinguish measured behavior from an explanation proposed for that behavior.

Articles and supporting materials

A publication may include a technical article, a more detailed paper, data, code, and supplementary notes. Supporting materials should identify the version of the article to which they belong. A prepared study protocol is not a completed experiment, and a proposed explanation is not an established account of a model’s architecture.

Corrections and licensing

Published articles will include the appropriate author or maintainer contact and a correction history when needed. Reuse permissions will be stated with each release. No blanket license is implied for third-party data, code, or figures.