Primary sourceshao__meng
View sourceEvent ·Multi-source
Anthropic 分享 Dreaming:让 Claude 在任务外离线整理 Agent 记忆
Anthropic 研究员 Lance Martin 分享了名为 dreaming 的后台记忆整理方法:不更新模型权重,而是在任务外回顾记录、纠错、去重并组织知识。在约 35 万份文件、4.2 亿 token 的内部语料实验中,单靠搜索的评分标准覆盖率为 53%,wiki 配合搜索提升到 61%,每题少用约 2 万 token。在 CL-Bench 数据库任务中,较强模型开始保存经过验证的规则和依据,只记失败的设定反而削弱收益。合同案例里,改为每份合同单独一页并逐字引用条款后,最终报告的关键事实覆盖从 41/57 提升到 57/57。
Current read
Anthropic 把 Agent 的记忆整理比作做梦:后台纠错、归纳规则、建 wiki,搜索覆盖率从 53% 涨到 61%,做 Agent 的都该看看细节。
11 sources64° AI heatLast updated 2026/11/7 08:03:56
Evidence
11 sourcesRelated source 1IT之家
View sourceRelated source 2Anthropic
View sourceRelated source 3elvis
View sourceRelated source 4Gorden Sun
View sourceRelated source 5kimmonismus
View sourceRelated source 6AI Will
View sourceRelated source 7SuperTechFans
View sourceRelated source 8Latent.Space
View sourceRelated source 9VnExpress Số hóa
View sourceRelated source 10Decoder
View sourceConflict check
Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.