Unverified50% confidenceFactExact time
In an Agent system, the vector store plays the role of long-term memory, compensating for the deficiencies of large models' limited context windows and their inability to remember historical conversations and enterprise-private knowledge.
1
Sources
50%
Confidence
Long-term
Relevance
9/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLarge models themselves do not have memory storage capability; the ability of conversational products to remember context is due to an external system maintaining the context.77% similarUnverifiedOverly aggressive context compression may cause the agent to lose memory, which in turn negatively impacts task quality.76% similarUnverified仅靠扩大上下文窗口无法解决Agent的长期记忆问题75% similarUnverifiedLangGraph's state management leans toward short-to-medium-term memory at the workflow level, and for long-term knowledge governance across sessions and agents it provides only infrastructure primitives rather than a complete solution.75% similarUnverified当前主流Agent系统本质上是在人类已产生的知识空间中进行高维插值，缺乏对尚未被提问问题的原创感知力73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/930609API
curl https://kongchang.com/api/v1/knowledge/claims/930609MCP
get_claim(id=930609)