English(EN) PolyRoute: A Twelve-Language and Code-Mixed Benchmark for Typed-Decision Routing, with Baselines and a Leaderboard

新的PolyRoute基准测试多语言和混合语AI路由能力

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-11 05:00

研究人员推出了PolyRoute，这是一个旨在评估类型化决策路由模型在十二种语言和混合语文本中性能的新基准。该基准包含300多条不同语言和书写系统的支持消息，以及一个三方部门路由决策协议。初步结果显示，虽然经过微调的多语言模型达到了96.6%的准确率，但传统的TF-IDF结合逻辑回归达到了91.9%。该基准还突出了在混合语文本上的性能差异，其中一个以英语为基础的模型略微优于多语言检查点。 AI

影响为多语言和混合语AI路由建立了新的评估标准，有望推动跨语言AI能力的提升。

排序理由该项目描述了一篇介绍基准和基线结果的新学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 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: 34 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该项目描述了一篇介绍基准和基线结果的新学术论文。[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.AI TIER_1 English(EN) · Mohammed Nayeem Khan, Sarun Kumar K S · 2026-11-11 05:00

PolyRoute：一个十二种语言和代码混合的类型化决策路由基准，包含基线和排行榜

arXiv:2611.13843v1 Announce Type: cross Abstract: Typed-decision encoders answer a question about a text in one forward pass, but their behaviour outside English is rarely measured on a fixed yardstick. We release PolyRoute: 322 native-language support messages in 12 languages an…

本文如何被排名

报道来源 [1]

PolyRoute：一个十二种语言和代码混合的类型化决策路由基准，包含基线和排行榜

相关实体

相关话题