BriefGPT.xyz
Apr, 2023
多任务学习在开放域对话系统中的实证研究
An Empirical Study of Multitask Learning to Improve Open Domain Dialogue Systems
HTML
PDF
Mehrdad Farahani, Richard Johansson
TL;DR
研究使用辅助任务对自回归模型进行改进，提高其在开放域对话生成中的长期语境处理和一致性维护，结果在小型和中型的GPT-2模型上表明引入新的辅助任务虽然有所改进，但仍有发展空间。
Abstract
autoregressive models
used to generate responses in
open-domain dialogue
systems often struggle to take long-term context into account and to maintain consistency over a dialogue. Previous research in
→