Learning Simpler Language Models with the Differential State Framework
Authors:
Preprint
Abstract
Learning useful information across long time lags is a critical and difficult problem for temporal neural models in tasks like language modeling. Existing architectures that address the issue are often complex and costly to train. The Differential State Framework (DSF) is a simple and high-performing design that unifies previously introduced gated neural models. DSF models maintain longer-term memory by learning to interpolate between a fast-changing data-driven representation and a slowly changing, implicitly stable state. This requires hardly any more parameters than a classical simple recurrent network. Within the framework, a new architecture is presented, the Delta-RNN. In language modeling at the word and character levels, the Delta-RNN outperforms popular complex architectures, such as the Long Short Term Memory (LSTM) and the Gated Recurrent Unit (GRU), and, when regularized, perform comparably to several state-of-the-art baselines. At the subword level, Delta-RNN's performance is comparable to that of complex gated architectures.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Hybrid computing using a neural network with dynamic external memoryNature, 2016
- Minimal gated unit for recurrent neural networksInternational Journal of Automation and Computing, 2016
- Parsing costs as predictors of reading difficulty: An evaluation using the Potsdam Sentence CorpusJournal of Eye Movement Research, 2008
- Expectation-based syntactic comprehensionCognition, 2008
- Noisy Time Series Prediction using Recurrent Neural Networks and Grammatical InferenceMachine Learning, 2001
- Computation of Conditional Probability Statistics by 8-Month-Old InfantsPsychological Science, 1998
- First-order versus second-order single-layer recurrent neural networksIEEE Transactions on Neural Networks, 1994
- Acceleration of Stochastic Approximation by AveragingSIAM Journal on Control and Optimization, 1992
- Learning and Extracting Finite State Automata with Second-Order Recurrent Neural NetworksNeural Computation, 1992