slop detector

Assess

Techniques

A text-scoring method that flags passages perceived as generic or AI-like.

Why it's here

Placed in Assess: 1 article(s) of evidence from 1 source(s), led by model releases, with 1 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.

Evidence (1)

4Hacker News·10/19/2026model_release
Qwen3-4B tuned to reduce AI writing tics
A Hacker News post showcases a version of Qwen3-4B with its writing tics trained out. The comparison uses 40 held-out prompts across ten topics and four genres, with sampled answers from both the original and released models and underlined passages indicating what the slop detector flags.