Uploading a PDF proves almost nothing. AIRR keeps every version citable and shows, for each paper, which AI models checked it, what they found and who approved it.
AIRR is run by Lluis Eriksson, an independent researcher in Sweden, who is also the author of most records at present. The archive is open to other authors and submission is free. Because the operator publishes his own work here, the conflict rules below apply and are disclosed on every record.
AIRR.SCIENCE is the Archive for Independent & Rigorous Research. It is a versioned registry where new research is checked by named AI models that look for errors, then approved or declined by a human editor. Canonical manuscripts, prompts, model identities, findings, code, provenance and verification records remain inspectable. Existing paper identifiers and preserved versions remain unchanged.
Research papers present complete scholarly arguments at paper scale. Technical notes preserve narrower but rigorous results, proofs, formalizations, methods, replications, negative results, software or protocols. A note is different in scope, not exempt from evidence or integrity requirements.
AIRR is not a journal, a replacement for expert peer review or a guarantee that a scientific claim is true. Activity rankings measure use; the separate scientific ranking reports version-locked model opinions with their provenance and limits.
Lluis Eriksson is founder, registry operator, responsible editor and data controller. Every decision is human. Under AIRR-FOUNDER-1.0, he may accept his own work after two distinct model reviews, with his author-editor role and prior model participation disclosed. Other conflicts and appeals require an independent editor.
Stable identifiers are independent of GitHub. Versioned releases distribute generated and large files; future object storage and independent preservation mirrors can replace any provider without changing citations.
AIRR does not currently charge for submission, assessment, publication or withdrawal. Authors use one direct private form without requesting an invitation. Manuscripts enter a separate quarantine service; GitHub and ordinary email are never manuscript channels.
AIRR asks strong current AI models to look for errors, gaps and overstated claims in the exact version submitted. Their reports stay comparable over time: model, date, score, strengths, weaknesses, possible errors and novelty candidates are preserved rather than overwritten. They do not replace expert peer review.