… In responding to the above problem, this paper proposes a new algorithm CMLA (Continual
Meta-Learning Algorithm) based on meta-learning. CMLA cannot only extract the key …

… Since our approach is grounded in the probabilistic formulation of meta-learning as … In
this work, we focus on model-agnostic meta-learning (MAML) [14], a gradient-based meta-learning …

… we learn a representation for continual learning that promotes future … by meta-learning the
representations offline on a meta-training dataset. At meta-test time, we initialize the continual …

… for offline continual learning [12, 22]… meta-learning algorithm for efficient, online continual
learning. We first propose a base algorithm for continual meta-learning referred to as Continual-…

… In this section, we review prior studies on online metalearning (OML) and continual meta-learning
(CML). These learning frameworks aim to incrementally improve a learning algorithm …

… We focus on Continual Meta-Learning (CML), which targets accumulating and exploiting
meta-knowledge on a sequence of non-iid tasks. The primary challenge is to strike a balance …

… the application of meta-learning to continual learning, where learners continually accumulate
experience over time to more rapidly acquire new knowledge or skills. Continual learning …

… , one can compose arbitrarily many continual meta learning methods. To show that this …
instances by adapting previous continual learning methods to this meta learning framework. …

… Meta-learning with MAML has previously been extended to … is a meta-learning for online
learning (MOLe) algorithm that … In contrast to prior multi-task and meta-learning methods, our …

… This paper focuses on continual meta-learning, where few-shot tasks are heterogeneous
and sequentially available. Recent works use a mixture model for meta-knowledge to deal with …