Sparse Meta Networks for Sequential Adaptation and its Application to Adaptive Language Modelling
This addresses the need for scalable and efficient learning in complex, changing environments, such as adaptive language modeling, though it appears incremental as it builds on existing meta-learning and fast-weight memory concepts.
The paper tackles the problem of slow and data-intensive training of deep neural networks in dynamic environments by introducing Sparse Meta Networks, a meta-learning approach that enables fast online sequential adaptation. It demonstrates strong performance, including on large-scale adaptive language modeling.
Training a deep neural network requires a large amount of single-task data and involves a long time-consuming optimization phase. This is not scalable to complex, realistic environments with new unexpected changes. Humans can perform fast incremental learning on the fly and memory systems in the brain play a critical role. We introduce Sparse Meta Networks -- a meta-learning approach to learn online sequential adaptation algorithms for deep neural networks, by using deep neural networks. We augment a deep neural network with a layer-specific fast-weight memory. The fast-weights are generated sparsely at each time step and accumulated incrementally through time providing a useful inductive bias for online continual adaptation. We demonstrate strong performance on a variety of sequential adaptation scenarios, from a simple online reinforcement learning to a large scale adaptive language modelling.