1 paper
Takyoung Kim, Jinseok Nam, Chandrayee Basu +5
Conversational agents powered by large language models (LLMs) with tool integration achieve strong performance on fixed task-oriented dialogue datasets but remain vulnerable to una…