已过期50% 置信事实精确时间
Current LLMs perform stateless inference based on context windows and do not retain memory of previous interactions unless content is explicitly passed in as context.
1
来源数
50%
置信度
中期 (~90 天)
时效性
2026/7/2
首次发现
有效期至：2026/9/30(已过期)
来源
相关事实
部分验证LLM每次推理是无状态的，受上下文窗口有限性限制，本身没有持久记忆74% 相似待验证LLM本质上是无状态的推理引擎，每次调用模型时所有记忆都必须以文本形式塞入上下文窗口72% 相似待验证LLM每次推理时仅处理当前输入的token序列，不保留跨会话的持久化状态（无状态性）70% 相似待验证The context window in LLMs determines how many conversation turns the model can remember when making modifications to previously generated content68% 相似待验证Using RAG does not require modifying the LLM itself—retrieved knowledge is placed into the question's context and fed to the model67% 相似
引用此条事实
Stable URI
https://kongchang.com/claim/51016API
curl https://kongchang.com/api/v1/knowledge/claims/51016MCP
get_claim(id=51016)