Coverage atlas
Critical AI across the social sciences
A credible critique journal for social-science AI research has to work across the social sciences, not just one corner. This atlas maps the journal’s domain-classifiable artifacts — the human-expert benchmarks, its own per-paper critiques, and the live selection queue — onto 8 social-science domains, and shows where the journal has not yet critiqued.
| Domain | Benchmarks | Critiques | Queue | Calibration profile |
|---|---|---|---|---|
| Management, IS & marketing | 3 | 5 | 12 | ✓ field standard |
| Communication & media | 3 | 3 | 7 | ✓ field standard |
| Political science | 8 | 0 | 0 | ✓ field standard |
| Economics & finance | 6 | 1 | 0 | ✓ field standard |
| Sociology | 6 | 0 | 0 | ✓ field standard |
| Psychology | 5 | 0 | 0 | ✓ field standard |
| Public policy & criminology | 5 | 0 | 0 | ✓ field standard |
| Education | 3 | 0 | 0 | ✓ field standard |
Green = the journal has critiqued in that domain. Amber rows are white space: domains with benchmarks and/or queued papers but no critique yet.
Where the journal is thin
The journal has a human-expert standard and queued candidates, but no critique yet, in Political science, Sociology, Psychology, Public policy & criminology, Education. Naming the gap is the point: the calibration metric already has a field standard ready for these domains, so the next critiques can be measured against them the moment they are written. The queue ranks the specific papers; this atlas says which domains to reach for.
Coverage re-derives from the benchmark corpus, the critique registry and the queue; machine-readable at /critique/api/coverage.