Skip-Thought Vectors

TLDR

The authors propose an unsupervised method to learn a generic, distributed sentence encoder that will be publicly released. They train an encoder‑decoder model on continuous book text to reconstruct surrounding sentences, expand the vocabulary to a million words, and evaluate the resulting vectors on eight downstream tasks. The resulting vectors map semantically and syntactically similar sentences to nearby points and achieve strong performance on multiple benchmark tasks.

Abstract

We describe an approach for unsupervised learning of a generic, distributed sentence encoder. Using the continuity of text from books, we train an encoder-decoder model that tries to reconstruct the surrounding sentences of an encoded passage. Sentences that share semantic and syntactic properties are thus mapped to similar vector representations. We next introduce a simple vocabulary expansion method to encode words that were not seen as part of training, allowing us to expand our vocabulary to a million words. After training our model, we extract and evaluate our vectors with linear models on 8 tasks: semantic relatedness, paraphrase detection, image-sentence ranking, question-type classification and 4 benchmark sentiment and subjectivity datasets. The end result is an off-the-shelf encoder that can produce highly generic sentence representations that are robust and perform well in practice. We will make our encoder publicly available.