Learning to Generate Reviews and Discovering Sentiment

We explore the properties of byte-level recurrent language models. When given sufficient amounts of capacity, training data, and compute time, the representations learned by these models include disentangled features corresponding to high-level concepts. Specifically, we find a single unit which performs sentiment analysis. These representations, learned in an unsupervised manner, achieve state of the art on the binary subset of the Stanford Sentiment Treebank. They are also very data efficient. When using only a handful of labeled examples, our approach matches the performance of strong baselines trained on full datasets. We also demonstrate the sentiment unit has a direct influence on the generative process of the model. Simply fixing its value to be positive or negative generates samples with the corresponding positive or negative sentiment.

PDF Abstract ICLR 2018 PDF ICLR 2018 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Sentiment Analysis SST-2 Binary classification bmLSTM Accuracy 91.8 # 46
Subjectivity Analysis SUBJ Byte mLSTM Accuracy 94.60 # 9

Methods


No methods listed for this paper. Add relevant methods here