State-of-the-art methods for Word Sense Disambiguation (WSD) combine two\ndifferent features: the power of pre-trained language models and a propagation\nmethod to extend the coverage of such models. This propagation is needed as\ncurrent sense-annotated corpora lack coverage of many instances in the\nunderlying sense inventory (usually WordNet). At the same time, unambiguous\nwords make for a large portion of all words in WordNet, while being poorly\ncovered in existing sense-annotated corpora. In this paper, we propose a simple\nmethod to provide annotations for most unambiguous words in a large corpus. We\nintroduce the UWA (Unambiguous Word Annotations) dataset and show how a\nstate-of-the-art propagation-based model can use it to extend the coverage and\nquality of its word sense embeddings by a significant margin, improving on its\noriginal results on WSD.\n