InfLoRA：面向持续学习的无干扰低秩适配
InfLoRA: Interference-Free Low-Rank Adaptation for Continual Learning
AI总结：
本文提出 InfLoRA，通过在预训练权重子空间中注入低秩参数并设计无干扰子空间，兼顾持续学习的稳定性与可塑性，在多个数据集上超越现有最优方法。
摘要：
持续学习要求模型按顺序学习多个任务。在持续学习中，模型应当具备保持其在旧任务上性能的能力（稳定性），以及持续适应新任务的能力（可塑性）。近年来，参数高效微调（PEFT）通过冻结预训练模型并注入少量可学习参数来适应下游任务，已在持续学习中获得越来越多的关注。尽管现有基于 PEFT 的持续学习方法相比不基于 PEFT 的方法表现出更优的性能，但其中大多数并未考虑如何消除新任务对旧任务的干扰，这阻碍了模型在稳定性与可塑性之间取得良好平衡。在本工作中，我们提出了一种用于持续学习的新型 PEFT 方法，称为无干扰低秩适配（InfLoRA）。InfLoRA 注入少量参数来对预训练权重进行重参数化，并表明微调这些注入参数等价于在一个子空间内微调预训练权重。此外，InfLoRA 对该子空间进行设计，以消除新任务对旧任务的干扰，从而在稳定性与可塑性之间实现良好平衡。实验结果表明，InfLoRA 在多个数据集上优于现有的最先进持续学习方法。
英文摘要：
Continual learning requires the model to learn multiple tasks sequentially. In continual learning, the model should possess the ability to maintain its performance on old tasks (stability) and the ability to adapt to new tasks continuously (plasticity). Recently, parameter-efficient fine-tuning (PEFT), which involves freezing a pre-trained model and injecting a small number of learnable parameters to adapt to downstream tasks, has gained increasing popularity in continual learning. Although existing continual learning methods based on PEFT have demonstrated superior performance compared to those not based on PEFT, most of them do not consider how to eliminate the interference of the new task on the old tasks, which inhibits the model from making a good trade-off between stability and plasticity. In this work, we propose a new PEFT method, called interference-free low-rank adaptation (InfLoRA), for continual learning. InfLoRA injects a small number of parameters to reparameterize the pre-trained weights and shows that fine-tuning these injected parameters is equivalent to fine-tuning the pre-trained weights within a subspace. Furthermore, InfLoRA designs this subspace to eliminate the interference of the new task on the old tasks, making a good trade-off between stability and plasticity. Experimental results show that InfLoRA outperforms existing state-of-the-art continual learning methods on multiple datasets.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
