Rainproof: An Umbrella To Shield Text Generators From Out-Of-Distribution Data. (arXiv:2212.09171v2 [cs.CL] UPDATED)

Implementing effective control mechanisms to ensure the proper functioning
and security of deployed NLP models, from translation to chatbots, is
essential. A key ingredient to ensure safe system behaviour is
Out-Of-Distribution (OOD) detection, which aims to detect whether an input
sample is statistically far from the training distribution. Although OOD
detection is a widely covered topic in classification tasks, most methods rely
on hidden features output by the encoder. In this work, we focus on leveraging
soft-probabilities in a black-box …

arxiv chatbots control data detection distribution nlp nlp models security text training translation