A reviewer database is a markdown file defining a pool of AI reviewer personas for a specific research field. Two databases are bundled: Computer Architecture and Machine Learning & AI. To review papers in a different field, generate a new database and upload it below.
To build a new database: use the prompt at
src/ai_paper_review/prompts/database_generation.md
(replace [FIELD NAME], paste into any LLM, get a ready-to-run YAML),
then run ai-paper-review-generate-db --config my_field_cfg.yaml --out my_field_db.md
and upload the result below.
Full spec: Database Format.
Source files on disk: {{ databases_dir }}
| Database | Reviewers | Path | |
|---|---|---|---|
| {{ db.label }} | {{ db.n_reviewers }} | {{ db.path }} |
View
{% if db.can_delete %}
{% else %}
default
{% endif %}
|
Upload a .md reviewer database file. The file is parsed on
upload; malformed databases are rejected with an error message.
For filename rules, naming conventions, and how labels in the table are derived, see the Filename handling section in Database Format.