Unverified50% confidenceTradeoffExact time
大语言模型每次对话独立，无法对整个代码库形成持久认知模型，导致不同会话生成的代码在架构风格、命名规范、错误处理策略上存在系统性不一致
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
UnverifiedIn multi-Agent collaborative development, large language models have a tendency to be "overly enthusiastic" and may overstep their responsibilities by modifying code outside their designated scope.81% similarVerifiedLarge language models are prone to forgetting early information or hallucinating during long conversations, which in programming manifests as fabricating non-existent APIs or generating logically incorrect code.80% similarUnverifiedThe semantic structure of code differs fundamentally from natural language, making it difficult for embedding models to accurately preserve information such as identifier naming, call relationships, and scope.79% similarUnverified大语言模型生成代码缺乏对系统整体架构的持续感知，随着项目规模扩大会出现上下文窗口溢出、模块耦合混乱、测试覆盖缺失等问题79% similarVerifiedLarge language models themselves are stateless; each API call is independent, and the model does not automatically remember the content of previous conversations.78% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/596317API
curl https://kongchang.com/api/v1/knowledge/claims/596317MCP
get_claim(id=596317)