Unverified50% confidenceFactExact time
Transformer 架构下模型在推理时仅处理当前输入的 token 序列，不存在跨请求的内部状态持久化
1
Sources
50%
Confidence
Long-term
Relevance
7/7/2026
First Seen
Sources
Why Does AI Take Such Bad Notes? A Deep Dive into the Limits and Trade-offs of LLM Memory
hackernewshackernews7/3/2026
Related Claims
VerifiedThe standard Transformer has no persistent state during inference; every call is a stateless forward computation, and the model itself has no memory capability.78% similarVerifiedLarge models are based on the Transformer architecture; their inference process is stateless, each request is an independent forward propagation computation, and the model itself retains no information between requests.75% similarUnverifiedThe Transformer model is a purely sequence-to-sequence mapping function whose parameters are completely static during inference, with no mechanism to actively probe the state of the external environment during inference.75% similarUnverifiedThe inference mechanism of the Transformer architecture determines that the model itself has no persistent memory capability, and all 'memory' relies entirely on the context provided in the current session.74% similarUnverifiedCurrent mainstream large language models adopt the autoregressive generation mechanism of the Transformer architecture, generating token by token from left to right, and cannot go back to modify already-output reasoning steps after generation.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/185818API
curl https://kongchang.com/api/v1/knowledge/claims/185818MCP
get_claim(id=185818)