Unverified50% confidenceFactExact time
This project implements long-term memory by storing key information in the file system to overcome the limitations of large language models' finite context length and the loss of information when a session ends.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/17/2026
First Seen
Valid until: 12/16/2026
Sources
Related Entities
Related Claims
VerifiedResearch shows that large language models exhibit a significant "Lost in the Middle" phenomenon when processing long texts, retaining information from the beginning and end of the input text well, but showing a marked decline in attention to information in the middle.75% similarUnverified陪伴类AI留存困境的技术根源在于当前主流大语言模型的长期记忆机制不成熟，对话窗口外的历史信息往往丢失75% similarUnverified当训练数据中某些文本出现频率极高时，大语言模型可能产生记忆化（Memorization）现象，能够近乎逐字输出特定段落73% similarVerifiedLarge language models themselves are stateless and do not persistently store conversational memory on the server side; memory must be explicitly maintained and passed by the caller.71% similarVerifiedLarge language models themselves have no persistent memory across sessions, and project context must be reloaded for each new task.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/930600API
curl https://kongchang.com/api/v1/knowledge/claims/930600MCP
get_claim(id=930600)