Unverified50% confidenceFactExact time
The output of a general-purpose LLM is an open-ended token sequence that requires autoregressive token-by-token generation, with computational cost growing linearly with output length.
1
Sources
50%
Confidence
Long-term
Relevance
9/24/2026
First Seen
Sources
Related Entities
Related Claims
Unverified现代主流LLM采用自回归生成方式，逐个token按顺序生成74% similarUnverified输出生成是严格串行的自回归过程，每生成一个token都需要一次完整的模型前向传播68% similarUnverifiedLLM自回归解码阶段每生成一个token都需从显存加载全量模型权重，批量较小时形成内存墙68% similarUnverifiedLLM的每次调用可以被建模为伯努利试验，N次独立调用的通过次数服从二项分布B(N, p)67% similarUnverifiedProduction-grade LLM applications often require multiple calls to the LLM, parsing of outputs, and passing context between different steps.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/947193API
curl https://kongchang.com/api/v1/knowledge/claims/947193MCP
get_claim(id=947193)