Fine-tuned on a MacBook Air, qwen3-4b-deslop cuts AI slop score from 121 to 12.5 per 1,000 words
lmoroney · x · 2026-10-20
Erik Kannike fine-tuned Qwen3-4B-Instruct on a MacBook Air to strip assistant writing habits (em dashes, tidy contrast lines, lists of three), releasing it as qwen3-4b-deslop. On 40 held-out prompts, slop score fell from 121 to 12.5 per 1,000 words, em dashes hit zero, and it still aced 30/30 factual questions.
Caveats: the voice is flat, it still invents names and places, and scoring uses a custom regex style guide with no blind reader test. The key insight is the task-fit filter: an earlier version scored lower but wrote opinion pieces with no opinion, so Kannike added checks for whether each answer did what the prompt asked. Check task fit before fussing over punctuation.
Related event: Developer Fine-tunes Qwen3-4B on a MacBook Air to Remove AI Writing Style(2 posts)→
More from Models
- Hidden "6.5" found in source code of OpenAI's new opensi.lol domain, countdown points to Nov 1 — kimmonismus · 2026-10-20
- Codex quota keeps getting reset: users say gpt-6.1-sol barely dents limits, want it to stay — bytebot · 2026-10-20
- Fudan & Shanghai AI Lab map PEFT recipes for MoE LLMs: tune experts alone, LoRA/DoRA win — kalyan_kpl · 2026-10-20
- Reddit User: New Models Have Mostly Stopped Hallucinating — Have LLMs Crossed the Reliability Line? — LiveNotWork · 2026-10-20
- 26-30B is the best dense model size range, argues ML commentator — auto_grad_ · 2026-10-20
- Researchers push back on Opus 4/4.1 deprecation, citing alignment and model welfare concerns — Sauers_ · 2026-10-20