提出无需存储旧数据的抗漂移特征空间，解决无样本持续学习中的遗忘问题。
LoRA Subtraction for Drift-Resistant Space in Exemplar-Free Continual Learning
- 通过减去旧任务LoRA权重构建抗漂移特征空间
- 在多个数据集上长序列任务中实现最优性能
- 参数高效且无需显式建模特征变化，适合资源受限场景
在持续学习中，灾难性遗忘常由特征漂移引发。这一问题在无样本持续学习（EFCL）中尤为突出，因无法保留旧任务样本，难以维持先验知识。现有方法依赖静态特征或过时统计量，无法捕捉特征空间的动态演化，导致性能随时间下降。本文提出抗漂移空间（DRS），无需显式特征建模或存储旧任务数据。引入一种新型参数高效的微调方法——低秩适配减法（LoRA-），在处理新任务前，从预训练权重中减去旧任务的LoRA权重，以构建用于训练的DRS。该方法提升稳定性、效率并简化实现。同时，通过三元组损失稳定特征漂移，增强模型可塑性。实验表明，该方法在多个数据集上均取得领先效果，尤其在长任务序列中表现优异。
原文摘要 · Abstract (English)
In continual learning (CL), catastrophic forgetting often arises due to feature drift. This challenge is particularly prominent in the exemplar-free continual learning (EFCL) setting, where samples from previous tasks cannot be retained, making it difficult to preserve prior knowledge. To address this issue, some EFCL methods aim to identify feature spaces that minimize the impact on previous tasks while accommodating new ones. However, they rely on static features or outdated statistics stored from old tasks, which prevents them from capturing the dynamic evolution of the feature space in CL, leading to performance degradation over time. In this paper, we introduce the Drift-Resistant Space (DRS), which effectively handles feature drifts without requiring explicit feature modeling or the storage of previous tasks. A novel parameter-efficient fine-tuning approach called Low-Rank Adaptation Subtraction (LoRA-) is proposed to develop the DRS. This method subtracts the LoRA weights of old tasks from the initial pre-trained weight before processing new task data to establish the DRS for model training. Therefore, LoRA- enhances stability, improves efficiency, and simplifies implementation. Furthermore, stabilizing feature drifts allows for better plasticity by learning with a triplet loss. Our method consistently achieves state-of-the-art results, especially for long task sequences, across multiple datasets.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。