Universal Sentence Encoder

TL;DR

This paper introduces efficient universal sentence encoding models designed for transfer learning in NLP, demonstrating superior performance over word-level methods, especially with limited training data, and addressing model bias detection.

Contribution

The paper presents novel sentence encoding models that balance accuracy and computational efficiency, outperforming traditional word-level transfer learning baselines.

Findings

Sentence embeddings outperform word-level transfer learning.

Good performance achieved with minimal supervised data.

Models show potential in bias detection tasks.

Abstract

We present models for encoding sentences into embedding vectors that specifically target transfer learning to other NLP tasks. The models are efficient and result in accurate performance on diverse transfer tasks. Two variants of the encoding models allow for trade-offs between accuracy and compute resources. For both variants, we investigate and report the relationship between model complexity, resource consumption, the availability of transfer task training data, and task performance. Comparisons are made with baselines that use word level transfer learning via pretrained word embeddings as well as baselines do not use any transfer learning. We find that transfer learning using sentence embeddings tends to outperform word level transfer. With transfer learning via sentence embeddings, we observe surprisingly good performance with minimal amounts of supervised training data for a transfer task. We obtain encouraging results on Word Embedding Association Tests (WEAT) targeted at detecting model bias. Our pre-trained sentence encoding models are made freely available for download and on TF Hub.