MediQ evaluates proactive information-seeking behavior in Large Language Models by simulating interactions between a Patient System and an Expert System to ensure reliable clinical reasoning under incomplete context.

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)

• Large Language Models concept