Scoring generations of the same news against each other to filter AI slop

ivan_bezdomny · x · 2026-08-25

Asked what dataset he started with to remove AI slop, ivanbezdomny explains his approach: he scores different generations of the same news against each other to measure quality (noting that SFT doesn't really work anyway), and uses real tweets to count word and phrase frequencies as a proxy for quality writing on the same topic. A longer post is in the works.

Related event: Developer Shares Recipe for De-Slopping AI Writing: Multi-Version Self-Scoring(3 posts)→

Original post →

More from Research

Research channel →