PulseAugur
实时 09:31:26
English(EN) xDailyBench: Benchmarking LLMs on Professional Consultation for Real-Life Problems

新的基准测试 xDailyBench 测试大型语言模型在现实世界专业咨询任务中的表现

研究人员推出 xDailyBench,这是一个旨在评估大型语言模型 (LLM) 在处理现实生活、专业咨询任务方面的能力的新基准测试。该基准测试包含 51 个场景中的 248 个任务,侧重于用户实际提出的请求,这些请求通常涉及从上下文中推断未明确说明的需求。对 11 个前沿模型的评估显示,尽管最佳模型的任务级别得分达到了 75.6%,但所有模型在处理隐含需求方面比处理明确需求时都困难得多,这凸显了这是需要改进的关键领域。 AI

影响 凸显了大型语言模型在满足现实世界中用户隐含需求方面的持续瓶颈,为未来模型开发提供指导。

排序理由 该集群包含一篇介绍大型语言模型新基准测试的研究论文。[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 →

新的基准测试 xDailyBench 测试大型语言模型在现实世界专业咨询任务中的表现

本文如何被排名

Signal score
13 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群包含一篇介绍大型语言模型新基准测试的研究论文。[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, product
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]

  1. arXiv cs.AI TIER_1 English(EN) · Yongchang Peng, Qingshui Gu, Liya Zhu, Ge Zhang, Duo Wang, Haodong Wang, Jingzhe Ding, Tianhao Yu, Letian Gao, Yongjie Zhong, Chaoxin Li, Zixin Su, Jinchao Tao, Xingyu Ma, Xin'ao Guo, Feng Tian, Shiyuan Dong, Xiaoyan He, Sen Liu, Xin Chen, Jiajun Li, Zej… ·

    xDailyBench:在专业咨询真实问题上对大型语言模型进行基准测试

    arXiv:2609.07784v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for everyday assistance, yet existing benchmarks only partially reflect the requests users naturally make in practice. Real-world requests are often open-ended, casually specified, …