Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context

Abstract

We know very little about how neural language models (LM) use prior linguistic context. In this paper, we investigate the role of context in an LSTM LM, through ablation studies. Specifically, we analyze the increase in perplexity when prior context words are shuffled, replaced, or dropped. On two standard datasets, Penn Treebank and WikiText-2, we find that the model is capable of using about 200 tokens of context on average, but sharply distinguishes nearby context (recent 50 tokens) from the distant history. The model is highly sensitive to the order of words within the most recent sentence, but ignores word order in the long-range context (beyond 50 tokens), suggesting the distant past is modeled only as a rough semantic field or topic. We further find that the neural caching model (Grave et al., 2017b) especially helps the LSTM to copy words from within this distant context. Overall, our analysis not only provides a better understanding of how neural LMs use their context, but also sheds light on recent success from cache-based models.

Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study

“…Since this work aims at investigating and gaining an understanding of the kinds of information a generative neural response model learns to use, the most relevant pieces of work are where sim- ilar analyses have been carried out to understand the behavior of neural models in other settings. An investigation into how LSTM based unconditional language models use available context was carried out by Khandelwal et al (2018). They empirically demonstrate that models are sensitive to perturbations only in the nearby context and typically use only about 150 words of context.…”

Post-Processing of Word Representations via Variance Normalization and Dynamic Embedding

“…Intuitively, words that are closer in a sentence should have stronger correlation. This has been verified in [6]. Thus, it is promising to design a new word embedding method that not only captures the context information but also models the dynamics in a word sequence.…”

An Augmented Transformer Architecture for Natural Language Generation Tasks

“…Although RNNs are widely adopted, there are still some drawbacks on RNNs: Back propagation procedure through time usually suffers from gradient vanishing and explosion [14]; The training process is also hard to be parallelized because of the consecutive operations. What's more, previous work has showed that LSTM is hard to tackle with longer sentences (say, more than 200 context words) [15], which hinders their further spreading.…”