Active Long Term Memory Networks

Abstract

Continual Learning in artificial neural networks suffers from interference and forgetting when different tasks are learned sequentially. This paper introduces the Active Long Term Memory Networks (A-LTM), a model of sequential multi-task deep learning that is able to maintain previously learned association between sensory input and behavioral output while acquiring knew knowledge. A-LTM exploits the non-convex nature of deep neural networks and actively maintains knowledge of previously learned, inactive tasks using a distillation loss. Distortions of the learned input-output map are penalized but hidden layers are free to transverse towards new local optima that are more favorable for the multi-task objective. We re-frame the McClelland's seminal Hippocampal theory with respect to Catastrophic Inference (CI) behavior exhibited by modern deep architectures trained with back-propagation and inhomogeneous sampling of latent factors across epochs. We present empirical results of non-trivial CI during continual learning in Deep Linear Networks trained on the same task, in Convolutional Neural Networks when the task shifts from predicting semantic to graphical factors and during domain adaptation from simple to complex environments. We present results of the A-LTM model's ability to maintain viewpoint recognition learned in the highly controlled iLab-20M dataset with 10 object categories and 88 camera viewpoints, while adapting to the unstructured domain of Imagenet with 1,000 object categories.

References in corpus (13)

• Distilling the Knowledge in a Neural Network
• Do Deep Nets Really Need to be Deep?
• Exact solutions to the nonlinear dynamics of learning in deep linear neural networks
• An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
• Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation
• Deep Convolutional Inverse Graphics Network
• Beyond Sharing Weights for Deep Domain Adaptation
• Actor-Mimic: Deep Multitask and Transfer Reinforcement Learning
• Bridging the Gaps Between Residual Learning, Recurrent Neural Networks and Visual Cortex
• Unifying distillation and privileged information
• A Deep Hierarchical Approach to Lifelong Learning in Minecraft
• Do Deep Convolutional Nets Really Need to be Deep and Convolutional?
• Blending LSTMs into CNNs

Cited by in corpus (9)

• Born Again Neural Networks
• Incremental Learning for Semantic Segmentation of Large-Scale Remote Sensing Data
• Lifelong Generative Modeling
• A Strategy for an Uncompromising Incremental Learner
• Continual Classification Learning Using Generative Models
• Continual Learning: Tackling Catastrophic Forgetting in Deep Neural Networks with Replay Processes
• Learning Translation Invariance in CNNs
• Closed-Loop Memory GAN for Continual Learning
• Growing a Brain: Fine-Tuning by Increasing Model Capacity