CogMem:可演化认知记忆架构用于长期对话
Get the story
2026年10月9日,arXiv Computation and Language 发布一手论文《CogMem:从检索到重构,构建面向长期对话的可演化认知记忆》。论文提出 CogMem 认知记忆架构,面向长期对话场景,核心为 PEC²F(Person-Event-Concept-Claim-Fact)图模式:Claim 节点用于保留主观陈述的来源与对象,Fact 与 Event 节点分别表达语义知识和情节知识。该架构将对话轮次增量转化为带溯源的图记录,并整合为高层事实,同时按时间范围调和冲突并进行更新。目前报道仅涉及该论文内容,未见后续实验数据、代码发布或第三方验证信息。
Generated from reports · updated 3 hr ago
Timeline
Follow the coverage from different angles.
- arXiv · Computation and LanguageCogMem:从检索到重构,构建面向长期对话的可演化认知记忆
论文提出 CogMem 认知记忆架构,基于 PEC²F(Person-Event-Concept-Claim-Fact)图模式,用 Claim 节点保留主观陈述的来源与对象,Fact 与 Event 节点表达语义和情节知识,并把对话轮次增量转为带溯源的图记录、整合为高层事实、按时间范围调和冲突更新。
Heat trend
Current heat 9·Comparable peak 10(Oct 9)·Comparable change over 24 hours –
The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.