Reference check

feature · paper open

Declaration 8fa4322f9a40 has not been accepted by anyone. · no paper read under it

Do the cited references resolve, and to what?

A capability that revises existing artifacts. It runs once, and then it is part of the schema.

Part of Tree — What does the paper assert, and how do its claims depend on each other?

How it works

Two different operations under one name, and the difference is the whole of what this layer is worth. Where a role: literature-context claim already carries a top-level doi:, the layer confirms with CrossRef that it resolves to a real paper: that is the anti-hallucination check. Where there is no DOI, it extracts hints from the claim text and slug, queries CrossRef, and writes back the highest-scoring match above a threshold — a search, which can return a confident match to the wrong paper.

The report is what the layer produces; the confirmed DOIs land in the claim files themselves. Two earlier implementations both edited those files and disagreed about the result, and one of them recorded a fuzzy title match as a confirmed DOI.

How to run it, in the reference

Rests on

Feeds

Nothing — this is a leaf.

How it is defined

What this layer reads besides its dependencies. Each is a declared input: its content is hashed into every run, so editing one makes those runs stale.

extract/claim_graphs/verify_refs.pythe script that runs itnot in the repository

The declaration names this path and the repository does not have it. An input that does not exist hashes to nothing, so it cannot make a run stale — the layer is declared to depend on something it is not in fact tracking.

What it produces3 results

One paper, as the worked example — Ejdrup, v1. Read from runs/ejdrup-2026-dopamine/reference-check.output.json · 1 KB. resolution_pct 100

# paper_slugclaim_slugstatusdoititleauthorsscorenote
1 ejdrup-2026-dopamine index confirmed 10.7554/eLife.105214 Computational modelling identifies key determinants of subregion-specific dopamine dynamics in the striatum Aske Ejdrup et al. —
2 ejdrup-2026-dopamine interprets-cragg-rice-vmax-ratio confirmed 10.1016/j.tins.2004.03.011 DAncing past the DAT at a DA synapse Stephanie J. Cragg, Margaret E. Rice — existing DOI resolves via CrossRef
3 ejdrup-2026-dopamine interprets-may-wightman-1989-fscv confirmed 10.1016/0006-8993(89)90835-4 Heterogeneity of stimulated dopamine overflow within rat striatum as observed with in vivo voltammetry Leslie J. May, R. Mark Wightman — existing DOI resolves via CrossRef

Across the corpus

8 not run · 2 stale·a paper links to its own cell, where this layer's output for it is rendered

Inputs and outputs

Reads, besides its dependencies
Produces
  • runs/{paper}/reference-check.output.json

One per paper — the table above links each one that exists.

Views
  • table — rendered above, over the 3 results in the artifact

Running it

The command comes from the declaration, so this text and what actually runs cannot diverge. pipeline.py run also runs the unmet dependencies first.

python3 scripts/pipeline.py run <paper> reference-check

Underneath, that runs cd extract && python3 -m claim_graphs.cli verify-refs --paper {paper}.