Unverified50% confidenceFactExact time
Traditional LLMs are passive text generators, whereas agents build upon this by adding a complete closed-loop capability of perception, planning, and execution
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
Unverified大语言模型的执行层通过 Function Calling（工具调用）机制将 LLM 的文本输出转化为对真实系统的操作指令71% similarUnverifiedThe large language model (LLM) is the core 'brain' of an AI Agent, but a standalone LLM can only generate text and cannot proactively execute tasks.70% similarUnverified在文本摘要、内容创作、代码生成、语义分类等带有生成或判断属性的任务上，LLM是无可替代的70% similarVerifiedThe root cause of prompt injection lies in the fact that LLM architecture encodes system prompts, user inputs, and externally retrieved content uniformly into the same continuous token sequence, making it impossible for the model to distinguish trusted instructions from untrusted data at the semantic level.70% similarUnverifiedA large language model (LLM) is a stateless text generation engine that accepts input and produces output, but does not possess the ability to take continuous actions.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/895659API
curl https://kongchang.com/api/v1/knowledge/claims/895659MCP
get_claim(id=895659)