arXiv cs.CL
· Papers
Towards Detecting Inconsistencies in End-to-end Generated TODs
arXiv:2607.09338v1 Announce Type: new Abstract: Generative AI is profoundly transforming the core technologies behind conversational systems, shifting from component-based to end-to-end approaches. However, Large Language Models (LLMs) may still generate inconsistencies, a critical issue particularly in Task-Oriented D