The 'LLM-as-a-Judge' paradigm serves as a scalable alternative to human expert annotation for evaluating medical LLMs.

Authors

• Person: Not available Organization: arXiv
  A Comprehensive Benchmark and Evaluation Framework for Multi ...

Sources

• A Comprehensive Benchmark and Evaluation Framework for Multi ... arxiv.org arXiv via serper

Referenced by nodes (1)

• LLM-as-a-judge concept