Concepts Fondamentaux
AI Agent
An LLM-powered system that autonomously takes actions in pursuit of a goal.
An AI agent uses an LLM as its reasoning core but can call external tools, browse the web, write code, and iterate over multiple steps to accomplish a goal. Agents loop: observe → reason → act → observe. They multiply token consumption significantly because each step involves model calls, tool outputs injected back as context, and often verification steps.
Termes Associés
Tool Use (Function Calling)
The ability for a model to call external functions or APIs to retrieve data or take actions.
Prompt Injection
An attack where malicious text in the environment overrides a model's instructions.
Context Window
The maximum number of tokens a model can process in a single request.
Grounding
Connecting model responses to verified, real-world information sources.