An Empirical Study of Adequate Vision Span for Attention-Based Neural Machine Translation
This work addresses computational bottlenecks for researchers and practitioners in machine translation, though it is incremental as it builds on existing attention models.
The paper tackled the computational inefficiency of attention mechanisms in neural machine translation by proposing a novel framework that dynamically reduces redundant score computations, achieving over 50% reduction in average window size with modest accuracy loss on English-Japanese and German-English tasks.
Recently, the attention mechanism plays a key role to achieve high performance for Neural Machine Translation models. However, as it computes a score function for the encoder states in all positions at each decoding step, the attention model greatly increases the computational complexity. In this paper, we investigate the adequate vision span of attention models in the context of machine translation, by proposing a novel attention framework that is capable of reducing redundant score computation dynamically. The term "vision span" means a window of the encoder states considered by the attention model in one step. In our experiments, we found that the average window size of vision span can be reduced by over 50% with modest loss in accuracy on English-Japanese and German-English translation tasks.% This results indicate that the conventional attention mechanism performs a significant amount of redundant computation.