Reflecting on the State of Rehearsal-free Continual Learning with Pretrained Models.
Lukas ThedeKarsten RothOlivier J. HénaffMatthias BethgeZeynep AkataPublished in: CoRR (2024)
Keyphrases
- learning tasks
- learning models
- learning process
- learning scenarios
- accurate models
- dynamic bayesian networks
- learning scheme
- neural nets
- online learning
- supervised learning
- reinforcement learning
- learning algorithm
- knowledge acquisition
- complex systems
- multi agent systems
- knowledge base
- finite state machines
- neural network
- qualitative models
- elementary school
- structured prediction
- data sets