阿里通义开源WebSailor 检索性能超DeepSeek R1、Grok-3等模型

智通财经
Jul 07, 2025

智通财经APP获悉,近日,阿里通义开源了网络智能体WebSailor,该智能体具备强大的推理和检索能力,在高难度智能体评测集BrowseComp上,WebSailor的成绩超越了DeepSeek R1、Grok-3等模型和智能体,一举登顶开源网络智能体榜单。目前WebSailor的构建方案及部分数据集已在Github开源。

为了让WebSailor更好地掌握复杂网页信息处理能力,通义团队设计了一套创新性的训练方法,包括三个关键模块:一是“地狱级试炼场”SailorFog-QA,通过真实网页构建图谱,制造信息混淆,让模型跨越多个页面整合线索,挑战人类认知极限;二是“重构推理逻辑”,摒弃冗长重复的推理链,让模型学习简洁、直击重点的思考方式,提升思维灵活性;三是“强化学习DUPO算法”,通过动态筛选高质量训练样本,提高训练效率2~3倍。

在权威评测平台 BrowseComp-en / BrowseComp-zh 中:WebSailor-72B 得分高居开源榜首;中文榜单中,与豆包(Doubao-Search)不分上下;更在英文榜单中 超过 Grok-3 等闭源模型。不仅如此,它在相对简单任务(如SimpleQA)中也表现优异。

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Most Discussed

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10