CULT: Continual Unsupervised Learning with Typicality-Based Environment Detection
Not on the map. Why?
Abstract (arXiv)
We introduce CULT (Continual Unsupervised Representation Learning with Typicality-Based Environment Detection), a new algorithm for continual unsupervised learning with variational auto-encoders. CULT uses a simple typicality metric in the latent space of a VAE to detect distributional shifts in the environment, which is used in conjunction with generative replay and an auxiliary environmental classifier to limit catastrophic forgetting in unsupervised representation learning. In our experiments, CULT significantly outperforms baseline continual unsupervised learning approaches. Code for this paper can be found here: https://github.com/oliveradk/cult
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 13 of its 16 references that are in the map. First posted on arXiv 17 Jul 2022.
Builds on
13 of 16 references.
References · 13
- 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
- 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
- 2013Auto-Encoding Variational Bayescited by 3,980
- 2013Playing Atari with Deep Reinforcement Learningcited by 3,211
- 2016Progressive Neural Networkscited by 1,332
- 2017Continual Learning with Deep Generative Replaycited by 1,152
- 2018Deep Anomaly Detection with Outlier Exposurecited by 939
- 2018Do Deep Generative Models Know What They Don't Know?cited by 443
- 2017A Conceptual Introduction to Hamiltonian Monte Carlocited by 207
- 2019Continual Unsupervised Representation Learningcited by 147
- 2018Life-Long Disentangled Representation Learning with Cross-Domain Latent Homologiescited by 43
- 2016Neurogenesis Deep Learningcited by 32
1 more
Built on by
Not cited yet.