Meituan Open-Sources LoHoSearch: Automatically Constructing High-Difficulty Search Agent Benchmarks with Knowledge Graphs
Original title:下一代搜索智能体评测基准!美团开源LoHoSearch,用知识图谱校准AI能力认知
AI summary
The Meituan LongCat team open-sourced LoHoSearch, a search agent evaluation benchmark that automatically generates questions from a knowledge graph covering 7.62 million Wikipedia entities, and includes 544 manually verified questions spanning 11 domains.
Key Takeaways
- The Meituan LongCat team open-sourced LoHoSearch
- a search agent evaluation benchmark that automatically generates questions from a knowledge graph…
- 62 million Wikipedia entities,and includes 544 manually verified questions spanning 11 domains.
Quick Facts
- Published
- Updated
- Source
- 公众号:龙猫LongCat(美团)
- Category
- Papers & Research
- AI score
- 67
LoHoSearch 用知识图谱自动生成高难度搜索题目,让当前最强模型准确率仅三成多,常用上下文管理策略在此几乎失效,为评测和优化长程搜索智能体提供了更苛刻也更有用的试验场。
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Meituan Open-Sources LoHoSearch: Automatically Constructing High-Difficulty Search Agent Benchmarks with Knowledge Graphs. July 23, 2026. https://new.nat6.net/en/news/cmsdu2t2a011grofzjokgp6gn/