#interactive systems

Reevaluating Uncertainty in Interactive LLM Agents

A recent paper critiques the classical uncertainty framework for large language model agents, suggesting a need for a new approach to better handle clarification seeking.