Overcoming catastrophic forgetting in neural networks
This addresses a crucial limitation for AI systems that need to learn tasks sequentially, representing a significant advance rather than an incremental improvement.
The paper tackled the problem of catastrophic forgetting in neural networks, showing that it is possible to train networks to maintain expertise on old tasks by selectively slowing down learning on important weights, and demonstrated effectiveness on MNIST classification and Atari 2600 games.
The ability to learn tasks in a sequential fashion is crucial to the development of artificial intelligence. Neural networks are not, in general, capable of this and it has been widely thought that catastrophic forgetting is an inevitable feature of connectionist models. We show that it is possible to overcome this limitation and train networks that can maintain expertise on tasks which they have not experienced for a long time. Our approach remembers old tasks by selectively slowing down learning on the weights important for those tasks. We demonstrate our approach is scalable and effective by solving a set of classification tasks based on the MNIST hand written digit dataset and by learning several Atari 2600 games sequentially.