Jump go di content
Glossary

Hallucination

Hallucination na confident but unsupported or fabricated output from language model: claim wey sound plausible but no get basis in provided evidence or reality. Hallucinations na central risk in automating knowledge work, and grounding with cited evidence na primary mitigation.

Synonym dem: AI hallucination, fabrication, confabulation, ungrounded output

Hallucination na wetin happen when fluency outrun truth. Because model optimize for plausible continuations, e fit produce specific names, numbers, or citations wey never real. For low-stakes drafting na nuisance; for governed work automation na hard failure wey fit mislead decision or trigger wrong action. Defenses dey architectural: retrieve real evidence, constrain generation to am, attach citations, and design explicit path for declining or escalating when evidence no support answer.

Question dem wey people dey ask well-well

Why language models hallucinate?
Models predict likely text, no be verified facts. Without retrieved evidence to constrain dem, dem fill gaps with statistically plausible but unverified statements.
How you reduce hallucination?
Ground answers in retrieved sources, require citations, verify claims against evidence, and route low-confidence or unsupported cases to person instead of returning guess.