A generic notion of a learning task is defined[^1] as $\mathcal{T}=\{\mathcal{L}(\mathbf{x}_1, \mathbf{a}_1, \ldots, \mathbf{x}_H, \mathbf{a}_H), q(\mathbf{x}_{1}), q(\mathbf{x}_{+1}|\mathbf{x}_t,\mathbf{a}_t),H\}$ with a loss function $\mathcal{L}$, a distribution over initial observations $q(\mathbf{x}_1)$, a transition distribution $q(\mathbf{x}_{t+1}|\mathbf{x}_t, \mathbf{a}_t)$ and an episode length $H$ (equal to 1 in i.i.d supervised problems). The model $f(\mathbf{x})=\mathbf{a}$ may generate samples of length $H$ by choosing an output $\mathbf{a}_t$ at each time $t$. [^1]: [Model-Agnostic Meta-Learning (Finn et al.)](https://arxiv.org/abs/1703.03400) In the multi-task meta-learning scenario, the model should adapt to a distribution over tasks $p(\mathcal{T})$, and it is trained (this is the phase called *meta-training*) to learn a new task $\mathcal{T_i}$ drawned from $p(\mathcal{T})$ from only $K$ samples drawn from $q_i$ and feedback $\mathcal{L}_{\mathcal{T_i}}$ generated by $\mathcal{T_i}$. It is then tested on new samples from $\mathcal{T_i}$; this is called *meta-testing*. Thus, the test error on sampled tasks $\mathcal{T_i}$ for the model $f$ serves as the training error of the meta-learning process.