Energy-Based Models for Continual Learning
arXiv:2011.12216
Abstract
We motivate Energy-Based Models (EBMs) as a promising model class for continual learning problems. Instead of tackling continual learning via the use of external memory, growing models, or regularization, EBMs change the underlying training objective to cause less interference with previously learned information. Our proposed version of EBMs for continual learning is simple, efficient, and outperforms baseline methods by a large margin on several benchmarks. Moreover, our proposed contrastive divergence-based training objective can be combined with other continual learning methods, resulting in substantial boosts in their performance. We further show that EBMs are adaptable to a more general continual learning setting where the data distribution changes without the notion of explicitly delineated tasks. These observations point towards EBMs as a useful building block for future continual learning methods.
arXiv commentProject page: https://energy-based-model.github.io/Energy-Based-Models-for-Continual-Learning
References in corpus (22)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Overcoming catastrophic forgetting in neural networks
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- Continual Learning Through Synaptic Intelligence
- PathNet: Evolution Channels Gradient Descent in Super Neural Networks
- Three scenarios for continual learning
- Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization
- Overcoming catastrophic forgetting with hard attention to the task
- Lifelong Learning with Dynamically Expandable Networks
- Towards Robust Evaluations of Continual Learning
- A Theory of Generative ConvNet
- Implicit Generation and Generalization in Energy-Based Models
- Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One
- Task Agnostic Continual Learning Using Online Variational Bayes
- Deep Value Networks Learn to Evaluate and Iteratively Refine Structured Outputs
- Meta-Learning Deep Energy-Based Memory Models
- Self-Supervised Learning Aided Class-Incremental Lifelong Learning
- ENGINE: Energy-Based Inference Networks for Non-Autoregressive Machine Translation
- Compositional Visual Generation and Inference with Energy Based Models
- Distributional Reinforcement Learning for Energy-Based Sequential Models
- Model Based Planning with Energy Based Models
- Benchmarking Approximate Inference Methods for Neural Structured Prediction