Questions
Common questions.
What is Hallucination?
A hallucination occurs when an AI model generates information that sounds plausible but is factually incorrect or fabricated.
How does Hallucination work?
In agentic systems, hallucinated tool calls, data references, or decisions can trigger real-world consequences.
Which terms are related to Hallucination?
Closely related concepts include Context Recall, Model Card, Agent Grounding, Model Distillation. Each is defined in the Prefactor glossary.
Where can I read more about Hallucination in practice?
The guide "Building agent hallucination guardrails" covers the applied side in depth.
Keep reading