Sparse Attentive Backtracking: Long-Range Credit Assignment in Recurrent Networks
arXiv:1711.02326
Abstract
A major drawback of backpropagation through time (BPTT) is the difficulty of learning long-term dependencies, coming from having to propagate credit information backwards through every single step of the forward computation. This makes BPTT both computationally impractical and biologically implausible. For this reason, full backpropagation through time is rarely used on long sequences, and truncated backpropagation through time is used as a heuristic. However, this usually leads to biased estimates of the gradient in which longer term dependencies are ignored. Addressing this issue, we propose an alternative algorithm, Sparse Attentive Backtracking, which might also be related to principles used by brains to learn long-term dependencies. Sparse Attentive Backtracking learns an attention mechanism over the hidden states of the past and selectively backpropagates through paths with high attention weights. This allows the model to learn long term dependencies while only backtracking for a small number of time steps, not just from the recent past but also from attended relevant past states.
References in corpus (9)
- Adam: A Method for Stochastic Optimization
- Densely Connected Convolutional Networks
- cuDNN: Efficient Primitives for Deep Learning
- Long Short-Term Memory Based Recurrent Neural Network Architectures for Large Vocabulary Speech Recognition
- Asymmetric LSH (ALSH) for Sublinear Time Maximum Inner Product Search (MIPS)
- EESEN: End-to-End Speech Recognition using Deep RNN Models and WFST-based Decoding
- Decoupled Neural Interfaces using Synthetic Gradients
- Knowing When to Look: Adaptive Attention via A Visual Sentinel for Image Captioning
- Training recurrent networks online without backtracking