# Preprint finds commit-first LLM judges can inherit model errors

_Published Wednesday, September 2, 2026 at 3:06 PM EDT · Science · Latest · Tier 2 — Notable_

A preprint auditing 24 default LLM-judge configurations across eight evaluation frameworks found none used commit-first judging. In controlled tests, a best-of-N search produced candidates accepted by a documented judge configuration despite failing held-out tests; commit-first removed that effect on one task but worsened results on another when the judge's committed answer was wrong. The authors also report errors in five of their own fifteen source-based claims.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2609.00088)

---
Canonical: https://techandbusiness.org/newswire/uS79b3LFgmbGdokm86sipz
Published: 2026-09-02T19:06:57.394Z
Story chronology: 2026-09-02T04:00:00.000Z
Retrieved: 2026-10-18T06:13:11.413Z
Publisher: Tech & Business (techandbusiness.org)