Existing benchmarks for Large Language Models (LLMs) fail to assess an LLM's ability to conduct structured consultations, manage dialogue flow, or exhibit safety behaviors during information gathering, despite their ability to evaluate domain knowledge retention.

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