Automatically Exposing Problems with Neural Dialog Models
Automatically Exposing Problems with Neural Dialog Models
复制标题
DOI：
10.18653/v1/2021.emnlp-main.37
复制
发表时间：
2021-09
期刊：
影响因子：
--
通讯作者：
Dian Yu;Kenji Sagae
中科院分区：
文献类型：
--
作者：
Dian Yu;Kenji Sagae
众所周知，神经对话模型存在生成不安全和不一致响应等问题。尽管这些问题至关重要且普遍存在，但它们大多是模型设计者通过交互手动识别的。最近，一些研究指示众包工作者刺激机器人触发此类问题。然而，人类利用仇恨言论等表面线索，却掩盖了系统性问题。在本文中，我们提出了两种方法，包括强化学习，以自动触发对话模型来生成有问题的响应。我们通过最先进的对话模型展示了我们的方法在暴露安全和矛盾问题方面的效果。
Neural dialog models are known to suffer from problems such as generating unsafe and inconsistent responses. Even though these problems are crucial and prevalent, they are mostly manually identified by model designers through interactions. Recently, some research instructs crowdworkers to goad the bots into triggering such problems. However, humans leverage superficial clues such as hate speech, while leaving systematic problems undercover. In this paper, we propose two methods including reinforcement learning to automatically trigger a dialog model into generating problematic responses. We show the effect of our methods in exposing safety and contradiction issues with state-of-the-art dialog models.