arXiv:2508.13171 · 中英对照阅读
认知工作空间：面向大语言模型的主动记忆管理——功能性无限上下文的实证研究
Cognitive Workspace: Active Memory Management for LLMs An Empirical Study of Functional Infinite Context
中文速览
大模型即使能处理超长文本，也常常不知道该主动记住什么、何时调用旧信息，传统检索增强生成（RAG）更像是“问了才查”，难以维持持续工作的记忆。论文提出“认知工作空间”（Cognitive Workspace），借鉴人类工作记忆和外部笔记的用法，让系统主动筛选整理信息、保存可持续的工作状态，并按任务需要调整上下文。实验中，它平均有58.6%的记忆可以在后续任务中重复利用，而传统RAG为0%，整体效率提升17%至18%，多类任务上的差异也具有统计显著性。这个思路的重要之处在于，它把提升大模型能力的重点从“塞进更多文本”转向“更聪明地管理记忆”，为构建能跨任务持续思考的系统提供了新方向。
摘要
Large Language Models (LLMs) face fundamental limitations in context management despite recent advances extending context windows to millions of tokens. We propose Cognitive Workspace, a novel paradigm that transcends traditional Retrieval-Augmented Generation (RAG) by emulating human cognitive mechanisms of external memory use. Drawing from cognitive science foundations including Baddeley’s working memory model [1, 2], Clark’s extended mind thesis [3], and Hutchins’ distributed cognition framework [4], we demonstrate that current passive retrieval systems fail to capture the dynamic, task-driven nature of human memory management. Our analysis of 2024-2025 developments reveals that while techniques like Infini-attention [5] and StreamingLLM [6] achieve impressive context lengths, they lack the metacognitive awareness and active planning capabilities essential for true cognitive extension. Cognitive Workspace addresses these limitations through three core innovations: (1) active memory management with deliberate information curation, (2) hierarchical cognitive buffers enabling persistent working states, and (3) task-driven context optimization that dynamically adapts to cognitive demands.
术语表
- Large Language Models (LLMs)
- 大语言模型（LLM）
- Cognitive Workspace
- 认知工作空间
- Retrieval-Augmented Generation (RAG)
- 检索增强生成（RAG）
- working memory model
- 工作记忆模型
- extended mind thesis
- 延展心智论
- distributed cognition framework
- 分布式认知框架
- Infini-attention
- Infini-attention
- StreamingLLM
- StreamingLLM
- metacognitive awareness
- 元认知意识
- active memory management
- 主动记忆管理
- hierarchical cognitive buffers
- 分层认知缓冲区
- task-driven context optimization
- 任务驱动的上下文优化
- memory reuse rate
- 记忆复用率
- Cohen’s d
- Cohen’s d
- transformer architecture
- Transformer架构
- quadratic attention complexity
- 二次方注意力复杂度
- Flash Attention 3
- Flash Attention 3
- MInference
- MInference
- MemGPT
- MemGPT
- Hierarchical Memory Transformer
- 分层记忆Transformer
- Self-RAG
- Self-RAG
- CRAG
- CRAG
- Adaptive RAG
- Adaptive RAG
- functional infinite context
- 功能性无限上下文
- cognitive load theory
- 认知负荷理论
- central executive
- 中央执行系统
- phonological loop
- 语音回路
- visuospatial sketchpad
- 视空间模板
- episodic buffer
- 情景缓冲器
- embedded-processes model
- 嵌入式过程模型