Hallucination
A hallucination is model output that is fluent and confident but unsupported by its source material or false about the world.
Ji et al.'s survey separates intrinsic hallucinations, which contradict the source, from extrinsic ones, which can't be verified from it. For agents, a fabricated tool result or citation is the costly kind.
Measure it with grounded test cases and reference-checking judges, and reduce it with retrieval, citations and permission to say "I don't know".
Learn more on Kojen
Related terms
- LLM-as-a-judgeLLM-as-a-judge is the practice of using a language model to grade another model's outputs against a rubric or a reference answer, instead of a human grader..
- Trust score (Kojen)Kojen's trust score is a 0–100 roll-up of an evaluation; for comprehensive evaluations it weights accuracy 50%, hallucination 30% and bias 20%..
- Context engineeringContext engineering is designing everything that goes into a model's context window, instructions, tools, retrieved documents, memory and history, so an agent has the information it needs and little else..
Sources
Put the definitions to work.
Kojen evaluates, tunes and red-teams your agents with these methods built in: Wilson intervals, pass^k, LLM judges with visible components. Free to start.
Start training free