arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.06950cs.CL
收藏

使用神经对话模型自动暴露问题

Automatically Exposing Problems with Neural Dialog Models

University of California, Davis(加州大学戴维斯分校)

Dian Yu, Kenji Sagae

摘要:

已知神经对话模型遭受诸如生成不安全和不一致的响应的问题。尽管这些问题是关键和普遍的,但它们大多是由模型设计人员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,人类利用表面的线索,如仇恨言论,同时留下系统性的问题隐藏。在本文中,我们提出了两种方法,包括强化学习自动触发对话模型生成有问题的响应。我们展示了我们的方法在暴露安全和矛盾问题与国家的最先进的对话模型的效果。

英文摘要:

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.

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。

↑