arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2012.14124cs.CLcs.AI
收藏

基于人工反例的神经文本生成

Neural Text Generation with Artificial Negative Examples

Graduate School of Informatics, Kyoto University(京都大学信息学研究生院);Graduate School of Science and Engineering, Ehime University(爱媛大学理工学研究生院);Academic Center for Computing and Media Studies, Kyoto University(京都大学计算与媒体研究学术中心);Salesforce Research(Salesforce研究院);Microsoft Research(微软研究院)

Keisuke Shirai, Kazuma Hashimoto, Akiko Eriguchi, Takashi Ninomiya, Shinsuke Mori

摘要:

以给定输入(例如机器翻译和图像描述)为条件的神经文本生成模型通常通过目标文本的最大似然估计来训练。然而,经过训练的模型在推理时会遇到各种类型的错误。在本文中,我们建议通过在强化学习框架中训练文本生成模型来抑制任意类型的错误,在该框架中,我们使用一个可训练的奖励函数,该函数能够区分参考文献和包含目标类型错误的句子。我们通过人为地将目标错误注入到参考文献中来创建这样的负例。在实验中,我们专注于两种错误类型,在模型生成的文本中重复和丢弃的令牌。实验结果表明,该方法在两个机器翻译任务和两个图像描述任务上都能有效地抑制生成错误,并取得显著的效果。

英文摘要:

Neural text generation models conditioning on given input (e.g. machine translation and image captioning) are usually trained by maximum likelihood estimation of target text. However, the trained models suffer from various types of errors at inference time. In this paper, we propose to suppress an arbitrary type of errors by training the text generation model in a reinforcement learning framework, where we use a trainable reward function that is capable of discriminating between references and sentences containing the targeted type of errors. We create such negative examples by artificially injecting the targeted errors to the references. In experiments, we focus on two error types, repeated and dropped tokens in model-generated text. The experimental results show that our method can suppress the generation errors and achieve significant improvements on two machine translation and two image captioning tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑