Neural machine translation (MT) models achieve strong results across a variety of settings, but it is widely believed that they are highly sensitive to "noisy" inputs, such as spelling errors, abbreviations, and other formatting issues. In this paper, we revisit this insight in light of recent multilingual MT models and large language models (LLMs) applied to machine translation. Somewhat surprisingly, we show through controlled experiments that these models are far more robust to many kinds of noise than previous models, even when they perform similarly on clean data. This is notable because, even though LLMs have more parameters and more complex training processes than past models, none of the open ones we consider use any techniques specifically designed to encourage robustness. Next, we show that similar trends hold for social media translation experiments -- LLMs are more robust to social media text. We include an analysis of the circumstances in which source correction techniques can be used to mitigate the effects of noise. Altogether, we show that robustness to many types of noise has increased.

翻译：神经机器翻译（MT）模型在各种场景下均取得了优异表现，但普遍认为它们对“噪声”输入（如拼写错误、缩写及其他格式问题）高度敏感。本文基于近期多语言MT模型及应用于机器翻译的大语言模型（LLMs）重新审视这一观点。令人略感意外的是，我们通过受控实验证明，这些模型对多种噪声的鲁棒性远超以往模型，即便在干净数据上表现相似时亦如此。这一发现值得关注，原因在于：尽管LLMs参数更多、训练过程更复杂，但我们考察的所有开源模型均未采用任何专门设计来增强鲁棒性的技术。随后我们证明，社交媒体翻译实验也呈现相似趋势——LLMs对社交媒体文本具有更强的鲁棒性。我们分析了源文本纠错技术可用于缓解噪声影响的具体情境。综上，我们证实模型对多种类型噪声的鲁棒性已有所提升。