Fumbling in Babel: An Investigation into ChatGPT's Language Identification Ability

arXiv:2311.09696v2 Announce Type: replace
Abstract: ChatGPT has recently emerged as a powerful NLP tool that can carry out a variety of tasks. However, the range of languages ChatGPT can handle remains largely a mystery. To uncover which languages ChatGPT `knows', we investigate its language identification (LID) abilities. For this purpose, we compile Babel-670, a benchmark comprising 670 languages representing 24 language families spoken in five continents. Languages in Babel-670 run the gamut from the very high-resource to the very low-resource. …

abstract arxiv chatgpt cs.cl however identification investigation language languages nlp tasks tool type