Prompting in information extraction tasks faces hallucination issues, where models overconfidently label negative inputs as entities or relations.

Authors

• Person: Not available Organization: Frontiers
  Combining large language models with enterprise knowledge graphs

Sources

• Combining large language models with enterprise knowledge graphs www.frontiersin.org Frontiers via serper

Referenced by nodes (2)

• information extraction concept
• prompting strategies concept