Many small claims, all under active replication

Abstract

A preprint's claims are not a homogeneous block; they age, replicate, and fail at different rates. We argue that the natural unit of replication is the individual claim, and we encode that argument operationally: every numbered claim below carries a structured active_replication annotation naming the replicating team, the start timestamp, and an expected-completion date. From an instrumentation run of the rrxiv reference instance ($n{=}312$ preprint–replication pairs across 14 months), pre-registering a replication target on a claim shifts median completion forward by approximately six weeks against a matched unregistered baseline. The paper is therefore both a substantive measurement of registration's effect on replication latency, and the canonical worked example of the active-replication pattern: it self-references its own annotations as evidence.