Auxiliary Task Update Decomposition: The Good, The Bad and The Neutral
This addresses the challenge of leveraging out-of-distribution data effectively in multitask learning for tasks with small training sets, representing an incremental improvement over prior methods.
The paper tackles the problem of selecting auxiliary tasks in multitask learning by proposing a model-agnostic framework that decomposes auxiliary task gradients into helpful, damaging, or neutral directions for the primary task, and it consistently outperforms strong baselines in text and image classification tasks.
While deep learning has been very beneficial in data-rich settings, tasks with smaller training set often resort to pre-training or multitask learning to leverage data from other tasks. In this case, careful consideration is needed to select tasks and model parameterizations such that updates from the auxiliary tasks actually help the primary task. We seek to alleviate this burden by formulating a model-agnostic framework that performs fine-grained manipulation of the auxiliary task gradients. We propose to decompose auxiliary updates into directions which help, damage or leave the primary task loss unchanged. This allows weighting the update directions differently depending on their impact on the problem of interest. We present a novel and efficient algorithm for that purpose and show its advantage in practice. Our method leverages efficient automatic differentiation procedures and randomized singular value decomposition for scalability. We show that our framework is generic and encompasses some prior work as particular cases. Our approach consistently outperforms strong and widely used baselines when leveraging out-of-distribution data for Text and Image classification tasks.