Open coding models turned telegraphic: filler words down 25–50% across 12 models
thesephist · x · 2026-11-08
PatronusAI tested 12 open models across 20 coding tasks and found a striking stylistic shift: a year ago open models wrote "Let me start by exploring the repository"; now several respond in clipped, article-free shorthand like "Need inspect repo. Grep."
- Filler words fell 25–50% between releases across the benchmark
- But the authors caution that terser output isn't cheaper — the thread digs into the cost implications
A systematic, quantified observation of how open coding models' output style has evolved.
More from Models
- LiquidAI ships official NVFP4 quantized weights for LFM3-1.7B and VL-2B — JosephJacks_ · 2026-11-08
- Qwen-4-27B early checkpoint tops GameFeel Bench, best open-source result by far — teortaxesTex · 2026-11-08
- 'Rank bottleneck' means two very different things in linear algebra vs systems optimization — kalomaze · 2026-11-08
- Double-exponential fit suggests model time horizons could hit ~80 work-years by end of 2027 — scaling01 · 2026-11-08
- Speculation: Anthropic's recent releases are all distills, Opus 5.5 may be Model 2 Early — haider1 · 2026-11-08
- Stanford Prof Anshul Kundaje on Weight Space Editing: Replicated Across Models, Limits Still Unclear — anshulkundaje · 2026-11-08