Jev Gallery
← Back to collection

Docs / source reviewed

Hacker News Judge

Judge Hacker News comments and assemble a source-linked thread overview

How this project uses Jev

Input
Thread context and individual comment text, capped at 400 comments per thread with top-level reactions prioritized
Jev decides
Stance, substance, quotability and whether a comment poses a question
Code executes
Code assembles verdicts and statistics, selects original comments as quotes and persists results; PostgreSQL BM25 handles full-text search

Evidence and limitations

Jev does not generate the overview prose; templates and aggregates assemble it. Comment counts and text lengths are capped, so coverage is not the full discussion or public opinion. Batch and on-demand judging incur inference costs. Database and model workflows were not independently run.

This project has not been run independently here. Author-reported results are not independently verified results.

Original sources

Explore related use cases

Browse this category →

Learn the approach