Exploring Effective Information Utilization in Multi-Turn Topic-Driven Conversations
Published in arXiv preprint, 2022
Recommended citation: Li, J., He, B., & Mi, F. (2022). Exploring Effective Information Utilization in Multi-Turn Topic-Driven Conversations. arXiv preprint arXiv:2209.00250. https://arxiv.org/abs/2209.00250
In order to expand the information that Pretrained Language Models can utilize, we encode topic and dialogue history information using certain prompts with multiple channels of Fusion-in-Decoder (FiD) and explore the influence of three different channel settings.