发表机构
INAOE; The University of Texas at El Paso（国家天体物理、光学与电子学研究所; 德克萨斯大学埃尔帕索分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出TORA几何路由框架，利用LoRA低秩结构按相似性决定知识迁移或隔离，在15个基准上避免有害路由，提升兼容任务性能并保护远距任务。
AI 中文摘要
文本分类中的持续学习（CL）面临两个关键挑战：灾难性遗忘和跨序列任务的负迁移。参数高效微调（PEFT）方法（如LoRA）通过学习模型参数的低秩更新来实现高效适应。然而，这些紧凑表示通常被孤立地训练，限制了它们在相关任务间的复用。我们引入了任务导向秩适应（TORA），一种几何路由框架，利用LoRA适配器的低秩结构，基于结构相似性决定是从最兼容的专家迁移知识（增强）还是隔离新任务（屏蔽）。在15个多样的文本分类基准上评估，TORA始终避免有害的路由决策：兼容任务超过其孤立性能同时减少训练时间，结构上距离较远的任务受到保护免受干扰且精度无损失。凭借单一几何阈值且不依赖任务标识或预定义序列，TORA为序列文本分类系统中的动态适配器路由提供了一种简单有效的方法。
英文摘要
Continual learning (CL) in text classification faces two critical challenges: catastrophic forgetting and negative transfer across sequential tasks. Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA enable efficient adaptation by learning low-rank updates of the model parameters. However, these compact representations are normally trained in isolation, limiting their reuse across related tasks. We introduce Task-Oriented Rank Adaptation (TORA), a geometric routing framework that leverages the low-rank structure of LoRA adapters to decide whether to transfer knowledge from the most compatible expert (Boosting) or isolate the new task (Shielding) based on structural similarity. Evaluated across 15 diverse text classification benchmarks, TORA consistently avoids harmful routing decisions: compatible tasks exceed their isolated performance while reducing training time, and structurally distant tasks are protected from interference with no loss in accuracy. With a single geometric threshold and no reliance on task identities or predefined sequences, TORA provides a simple and effective approach for dynamic adapter routing in sequential text classification systems.
CommentsPreprint submitted to CIARP2026
