Can a model hallucinate while summarizing a supplied document?
Yes. It may add an unsupported detail, confuse speakers, or turn a possibility into a conclusion. Compare important summary claims with the original text.
Explore This Question
Search for additional explanations using these tools:
More FAQs in What causes AI hallucinations?
No. Errors can also arise from ambiguity, missing context, incorrect source material, faulty inference, or failure to use relevant evidence that is already available.
No. Retrieval can provide evidence, but it can return unsuitable material, and the model can misinterpret correct material. Both stages need evaluation.
Still have questions?
Browse all our FAQs or reach out to our support team