# VALIDATION judge: VALIDATION judge (styles)

> Living preprint · as of 2026-10-25 · map node X-392 (experiment) · interim record, not a final result.

<!-- lint:exempt:start -->
- [X-392] VALIDATION judge · VALIDATION judge (styles) · question: Succinct vs S+ (one 4-5 sentence paragraph) vs two paragraphs. · result: correctness equal; wrong-unwarned ON 0.12→0.23/0.25. · set UXMATRIX 60, judge Claude Haiku 478 · verdict: adopted (Succinct stays)

<!-- lint:exempt:end -->
Read chapter 14, Closing the research: what a small model needs: https://r4topunk.github.io/boar-research/research/closure.md

## Facts cited {#facts}

Resolve every [F-…] marker in https://r4topunk.github.io/boar-research/data/facts.json. Each line gives the claim, the number with its metric family, set and status, judge, n, seeds, CI, date and caveat.

<!-- lint:exempt:start -->
- [F-2060] VALIDATION judge (styles): Succinct vs S+ (one 4-5 sentence paragraph) vs two paragraphs. correctness equal; wrong-unwarned ON 0.12→0.23/0.25 (outcome-label share) · set UXMATRIX 60 (status UNKNOWN) · judge Claude Haiku 478 · n UNKNOWN · seeds UNKNOWN · CI not computed · date 2026-10-24 · caveat: criterion frozen: yes (addendum 5); arm verdict: Succinct stays

<!-- lint:exempt:end -->
---

Cite as: r4to, independent researcher. RatLab — A Boar Research (2026), research record 1.0, as of 2026-10-25. Text and data CC BY 4.0. Fact ids or a link must travel with a quote. How to cite: https://r4topunk.github.io/boar-research/cite/ · https://r4topunk.github.io/boar-research/CITATION.cff.

AI-reviewed, not clinically reviewed. Nothing here is medical advice.