Unverified50% confidenceOpinionExact time
Vector databases cannot distinguish old from new versions when storing key decision memories, because embedding vectors encode only semantic similarity, not the "supersedes" relationship.
1
Sources
50%
Confidence
Long-term
Relevance
10/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedVector databases cannot express structured relationships (such as the source user, session, and time of a memory), cannot perform precise filtering and transactional updates, and their similarity thresholds also lack interpretability.79% similarUnverifiedThe problem of memory drift almost never exists in traditional software, because what is written to the database is explicit field values, yet it occurs very easily in architectures that allow models to autonomously generate and update memories.74% similarUnverified向量数据库通常缺乏原生的行级权限支持，需要在相似度计算时附加元数据过滤条件72% similarUnverifiedAn expanded context window does not equate to improved memory capability, as the model is still unable to precisely retrieve and associate historical information the way a database can.71% similarUnverified现有向量数据库记忆方案缺乏时效性衰减机制，无法区分记忆是否被采纳并产生正向效果，也没有机制合并冗余或删除有害记忆71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1054146API
curl https://kongchang.com/api/v1/knowledge/claims/1054146MCP
get_claim(id=1054146)