Unverified60% confidenceFactExact time
Large language models are essentially stateless functions, and a model's memory depends entirely on the content within the context window.
2
Sources
60%
Confidence
Long-term
Relevance
8/26/2026
First Seen
Sources
Related Entities
Related Claims
Unverified大语言模型的所有记忆都存储在当前的上下文窗口（Context Window）中，开启新Thread后这些信息全部清空79% similarVerified大语言模型的推理过程本质上是内存带宽受限（memory-bandwidth bound）的任务，每生成一个Token都需要将模型的全部权重从存储中读取一遍76% similarUnverifiedThe "forgetfulness" of large language models stems from their stateless inference architecture: each time the API is called, the model only processes the token sequence passed in the current request and retains no internal state across requests.76% similarUnverifiedLarge language models themselves are stateless, and each request is an independent computation; so-called 'remembering' usually relies on two mechanisms: one is concatenating historical conversation into the current prompt (the context window), and the other is retrieval-augmented generation (RAG) through external storage (such as vector databases or file systems).75% similarUnverifiedThe inference process of large language models is often constrained by memory bandwidth rather than pure floating-point compute.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/804538API
curl https://kongchang.com/api/v1/knowledge/claims/804538MCP
get_claim(id=804538)