TAME: Task Agnostic Continual Learning using Multiple Experts

TL;DR

TAME introduces a task-agnostic continual learning method that detects task shifts automatically, switches between multiple expert networks online, and outperforms existing methods on benchmark datasets with comparable model size.

Contribution

The paper presents TAME, a novel task-agnostic continual learning algorithm that infers task changes from loss deviations and dynamically switches between experts using a trained selector network.

Findings

Outperforms previous task-agnostic methods on benchmarks.

Effectively detects task shifts without prior task identity knowledge.

Maintains competitive model size through online pruning.

Abstract

The goal of lifelong learning is to continuously learn from non-stationary distributions, where the non-stationarity is typically imposed by a sequence of distinct tasks. Prior works have mostly considered idealistic settings, where the identity of tasks is known at least at training. In this paper we focus on a fundamentally harder, so-called task-agnostic setting where the task identities are not known and the learning machine needs to infer them from the observations. Our algorithm, which we call TAME (Task-Agnostic continual learning using Multiple Experts), automatically detects the shift in data distributions and switches between task expert networks in an online manner. At training, the strategy for switching between tasks hinges on an extremely simple observation that for each new coming task there occurs a statistically-significant deviation in the value of the loss function that marks the onset of this new task. At inference, the switching between experts is governed by the selector network that forwards the test sample to its relevant expert network. The selector network is trained on a small subset of data drawn uniformly at random. We control the growth of the task expert networks as well as selector network by employing online pruning. Our experimental results show the efficacy of our approach on benchmark continual learning data sets, outperforming the previous task-agnostic methods and even the techniques that admit task identities at both training and testing, while at the same time using a comparable model size.