Humpty Dumpty: Controlling Word Meanings via Corpus Modifications
ieee symposium on security and privacy, 2020.
Word embeddings, i.e., low-dimensional vector representations such as GloVe and SGNS, encode word meaningu0027u0027 in the sense that distances between wordsu0027 vectors correspond to their semantic proximity. This enables transfer learning of semantics for a variety of natural language processing tasks. Word embeddings are typically tra...More
Full Text (Upload PDF)
PPT (Upload PPT)