Unverified50% confidenceFactExact time
大语言模型有两个致命缺点：知识存在时间差和范围局限导致会编造，以及完全被动不会主动行动
1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen
Sources
AI大神的秘密：读懂底层逻辑，比会用提示词更重要
bilibilibest大模型7/9/2026
Related Claims
UnverifiedCurrent large language model agents generally lack persistent memory when executing complex tasks, making it difficult to remember previous decisions and their outcomes.74% similarUnverified现有大语言模型基本停留在因果阶梯的第一层，无法真正执行干预和反事实的因果推理71% similarUnverifiedLeCun 批评自回归语言模型在系统2慢思考（推理与规划）上存在根本性局限，且对物理世界缺乏基础（grounding）71% similarVerifiedThe core challenge of internalizing knowledge into model parameters is the Catastrophic Forgetting problem, which is a difficulty not yet fully resolved in the field of continual learning.71% similarUnverified对齐悖论指若模型真实理解比其能输出的语言更精细，则基于语言输出的对齐评估可能无法真正捕捉模型内部状态的对齐程度69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/484982API
curl https://kongchang.com/api/v1/knowledge/claims/484982MCP
get_claim(id=484982)