SPARC：面向LLM鲁棒持续学习的子空间感知提示适配
SPARC: Subspace-Aware Prompt Adaptation for Robust Continual Learning in LLMs
AI总结：
SPARC通过PCA低维子空间提示微调并结合LoRA，以极低参数开销实现LLM持续学习中的高效任务适配与知识保持。
摘要：
我们提出SPARC，一个面向大语言模型（LLMs）的轻量级持续学习框架，通过在低维空间中进行提示微调来实现高效的任务适配。借助主成分分析（PCA），我们识别出训练数据的一个紧凑子空间。在该低维空间中优化提示能够提升训练效率，因为它将更新集中在最相关的特征上，同时减少计算开销。此外，由于模型内部结构保持不变，预训练所获得的大量知识被完整保留，确保在适配过程中不会损害先前学到的信息。我们的方法在任务增量式和领域增量式持续学习设置下均实现了高知识保持率，同时仅微调模型0.04%的参数。另外，通过集成LoRA，我们增强了对计算约束的适应性，允许在准确率与训练成本之间进行权衡。在SuperGLUE基准上的实验表明，我们基于PCA的提示微调与LoRA相结合，在仅使用模型1%参数的情况下，既保持了完整的知识保持率，又提升了准确率。这些结果使我们的方法成为LLM持续学习中一种可扩展且资源高效的解决方案。
英文摘要：
We propose SPARC, a lightweight continual learning framework for large language models (LLMs) that enables efficient task adaptation through prompt tuning in a lower-dimensional space. By leveraging principal component analysis (PCA), we identify a compact subspace of the training data. Optimizing prompts in this lower-dimensional space enhances training efficiency, as it focuses updates on the most relevant features while reducing computational overhead. Furthermore, since the model's internal structure remains unaltered, the extensive knowledge gained from pretraining is fully preserved, ensuring that previously learned information is not compromised during adaptation. Our method achieves high knowledge retention in both task-incremental and domain-incremental continual learning setups while fine-tuning only 0.04% of the model's parameters. Additionally, by integrating LoRA, we enhance adaptability to computational constraints, allowing for a tradeoff between accuracy and training cost. Experiments on the SuperGLUE benchmark demonstrate that our PCA-based prompt tuning combined with LoRA maintains full knowledge retention while improving accuracy, utilizing only 1% of the model's parameters. These results establish our approach as a scalable and resource-efficient solution for continual learning in LLMs.
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。
