Unverified85% confidenceFactTime unknown
MemPalace在LongMemEval基准测试上，不使用LLM的纯检索方案前5召回率达到96.6%
1
Sources
85%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026
Sources
MemPalace：给AI Agent装上本地长期记忆的开源工具
bilibiliCharaplay
Related Entities
Related Claims
UnverifiedMemPalace的另一个调优方案在LongMemEval未参与调参的问题上达到98.4%的召回率72% similarUnverified根据LongMEM Eval基准测试，AgentMemory的检索准确率达到95.2%，比同类项目Mem0高出27个百分点69% similarUnverifiedDeepSuite基准测试的误报率为44.9%，漏报率为1.2%63% similarUnverified美团LongCat推出Lohosearch搜索智能体基准，11个前沿模型测试中最佳得分仅34.74%，远低于人类约90%的表现63% similarUnverifiedGLM5.1等国产模型在HumanEval、MBPP等标准编程基准测试上的得分已接近甚至超越部分国际竞品61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/3684API
curl https://kongchang.com/api/v1/knowledge/claims/3684MCP
get_claim(id=3684)