Jun 2026
When the Database Fails: Prompting LLM Dialogue Agents for Safe Recovery in Task-Oriented Dialogue
A lightweight prompting-based recovery approach that improves robustness without retraining or additional model calls is studied, and human annotation shows substantial agreement while supporting the validity of the automatic commitment-safety metric.
Mohammad Alijanpour Shalmani, Alale Rezvani Boroujeni, Ji Yuan
· SIGDIAL Conferences · 0 citations