---
id: 20260702-T0-08
title: "RoPoLL：基于鲁棒污染模型的 LLM 评判团"
title_en: "RoPoLL: A Robust Panel of LLM Judges"
url: https://ai.daily.yangsir.net/daily/20260702-T0-08
issue_date: 2026-07-02
publish_date: 2026-07-01T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2606.30931
---

# RoPoLL：基于鲁棒污染模型的 LLM 评判团

针对大模型评估中常用的“LLM 评判团”方法，新研究提出了 RoPoLL（Robust Panel of LLM Judges）。研究在 Huber 污染模型下形式化了评判团的统计行为，旨在解决单一评判者不可靠及现有评判团统计特性理解不足的问题，提升了评估结论的鲁棒性。

## English Version

**RoPoLL: A Robust Panel of LLM Judges**

This paper introduces RoPoLL, a robust framework for Panels of LLM Evaluators (PoLL). It formalizes the statistical behavior of LLM juries under the Huber contamination model, addressing the lack of understanding in current consensus-based evaluation methods and improving robustness.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2606.30931)

**详情页**：https://ai.daily.yangsir.net/daily/20260702-T0-08

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*