Feb. 26, 2024, 5:48 a.m. | Armand Stricker, Patrick Paroubek

cs.CL updates on arXiv.org arxiv.org

arXiv:2402.15248v1 Announce Type: new
Abstract: During task-oriented dialogues (TODs), human users naturally introduce chitchat that is beyond the immediate scope of the task, interfering with the flow of the conversation. To address this issue without the need for expensive manual data creation, we use few-shot prompting with Llama-2-70B to enhance the MultiWOZ dataset with user backstories, a typical example of chitchat interference in TODs. We assess the impact of this addition by testing two models: one trained solely on TODs …

70b abstract arxiv beyond conversation cs.cl data data creation few-shot flow human interference issue llama prompting the conversation type

Artificial Intelligence – Bioinformatic Expert

@ University of Texas Medical Branch | Galveston, TX

Lead Developer (AI)

@ Cere Network | San Francisco, US

Research Engineer

@ Allora Labs | Remote

Ecosystem Manager

@ Allora Labs | Remote

Founding AI Engineer, Agents

@ Occam AI | New York

AI Engineer Intern, Agents

@ Occam AI | US