BriefGPT.xyz
Aug, 2018
检索与精炼：用于对话的改进序列生成模型
Retrieve and Refine: Improved Sequence Generation Models For Dialogue
HTML
PDF
Jason Weston, Emily Dinan, Alexander H. Miller
TL;DR
本研究发展了一种融合了两种不同方法的模型，能够检索出机器人回答，并将其作为附加语境进行精细调节，从而产生了更准确、更生动、更吸引人的回答。在最近的 CONVAI2 挑战评估中，该模型在人类评估中明显优于传统的检索和生成模型。
Abstract
sequence generation models
for
dialogue
are known to have several problems: they tend to produce short, generic sentences that are uninformative and unengaging.
→