通过KILO应对LLM中的分布偏移：面向持续适应的知识引导学习
Tackling Distribution Shift in LLM via KILO: Knowledge-Instructed Learning for Continual Adaptation
AI总结：
提出KILO框架，结合动态知识图谱与指令微调，在四个领域上优于基线，有效缓解LLM持续学习中的分布偏移与灾难性遗忘。
摘要：
大型语言模型（LLMs）在面对领域偏移时常常出现性能下降，这主要是由于灾难性遗忘所致。在这项工作中，我们提出了KILO（面向持续适应的知识引导学习），一种新颖的持续学习框架，该框架将动态知识图谱与指令微调相结合。通过在训练过程中利用检索到的领域特定知识作为指导，KILO增强了对新领域的适应性以及对先前获取知识的保持能力。我们在WikiText-103上预训练模型，并评估了在四个不同目标领域上的顺序适应：BioASQ、SciQ、TweetEval和MIND。我们的实验表明，在后向迁移、前向迁移、F1分数、保持率和训练效率方面，KILO始终优于强基线方法，包括持续微调、ERNIE 2.0和CPT。这些结果凸显了结合结构化知识检索和指令提示以克服持续学习场景中领域偏移挑战的有效性。
英文摘要：
Large Language Models (LLMs) often suffer from performance degradation when faced with domain shifts, primarily due to catastrophic forgetting. In this work, we propose KILO (Knowledge-Instructed Learning for Continual Adaptation), a novel continual learning framework that integrates dynamic knowledge graphs with instruction tuning. By leveraging retrieved domain-specific knowledge as guidance during training, KILO enhances both adaptability to new domains and retention of previously acquired knowledge. We pretrain our model on WikiText-103 and evaluate sequential adaptation across four diverse target domains: BioASQ, SciQ, TweetEval, and MIND. Our experiments demonstrate that KILO consistently outperforms strong baselines, including continual fine-tuning, ERNIE 2.0, and CPT, in terms of backward transfer, forward transfer, F1 score, retention rate, and training efficiency. These results highlight the effectiveness of combining structured knowledge retrieval and instruction prompting to overcome domain shift challenges in continual learning scenarios.
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。
