English(EN) Holdout Best-of-N: Unbiased Evaluation and Its Cost

新评估方法提供无偏AI模型评估

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-07 04:00

研究人员推出了一种名为“Holdout Best-of-N”的新型机器学习模型评估方法，旨在提供无偏评估。该方法解决了在评估中使用用于选择的分数时夸大模型性能的问题。通过采用使用新分数进行选择的策略，该方法确保在各种分数分布下进行无偏评估，特别是对于高斯分数，其最小最大风险为 $\sigma^2/\sqrt K$ 阶。该系统设计为计算高效，有界分数的循环评估在池大小上具有统一的 $O(K^{-1})$ 风险。 AI

影响引入了一种更可靠的AI模型评估方法，可能带来更准确的性能评估和更好的模型选择。

排序理由该集群包含一篇详细介绍机器学习模型新评估方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 15 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍机器学习模型新评估方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Shrey Shah, Yinheng Li · 2026-10-07 04:00

持异者最佳N：无偏评估及其成本

arXiv:2610.08719v1 Announce Type: new Abstract: Reusing the scores that select a Best-of-$N$ winner can overstate its expected reward. We study evaluation from a fixed matrix of $K$ independent scores per candidate for a policy that selects using $J$ fresh scores. A single estima…

本文如何被排名

报道来源 [1]

持异者最佳N：无偏评估及其成本

相关实体

相关话题