arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2407.16920cs.CL
收藏

Train-Attention:在持续知识学习中元学习关注位置

Train-Attention: Meta-Learning Where to Focus in Continual Knowledge Learning

Yonsei University(延世大学)

Yeongbin Seo, Dongha Lee, Jinyoung Yeo

AI总结:

针对大语言模型持续知识学习中统一权重更新导致的低效与遗忘问题,提出 TAALM 方法,通过元学习动态预测 token 重要性以优化知识更新,并构建新基准 LAMA-ckl,实验证明其达到最先进性能且可与既有方法协同。

摘要:

以往关于大语言模型(LLMs)中持续知识学习(CKL)的研究主要集中在正则化、架构修改和 rehearsal 技术等方法上,以缓解灾难性遗忘。然而,这些方法天真地继承了标准训练过程的低效性,不加区分地对所有 token 施加统一的权重,这可能导致不必要的参数更新并加剧遗忘。为了解决这些不足,我们提出了一种新颖的 CKL 方法,称为 Train-Attention-Augmented Language Model(TAALM),该方法通过根据 token 的有用性动态预测并施加权重来提高学习效率。该方法采用元学习框架来优化 token 重要性预测,促进有针对性的知识更新并最小化遗忘。此外,我们观察到现有基准没有清晰地展示学习与保留之间的权衡,因此我们提出了一个新的基准 LAMA-ckl 来解决这一问题。通过在新引入的和已建立的 CKL 基准上进行的实验,TAALM 证明了其在基线之上的最先进性能,并且在与先前的 CKL 方法集成时也表现出协同兼容性。

英文摘要:

Previous studies on continual knowledge learning (CKL) in large language models (LLMs) have predominantly focused on approaches such as regularization, architectural modifications, and rehearsal techniques to mitigate catastrophic forgetting. However, these methods naively inherit the inefficiencies of standard training procedures, indiscriminately applying uniform weight across all tokens, which can lead to unnecessary parameter updates and increased forgetting. To address these shortcomings, we propose a novel CKL approach termed Train-Attention-Augmented Language Model (TAALM), which enhances learning efficiency by dynamically predicting and applying weights to tokens based on their usefulness. This method employs a meta-learning framework that optimizes token importance predictions, facilitating targeted knowledge updates and minimizing forgetting. Also, we observe that existing benchmarks do not clearly exhibit the trade-off between learning and retaining, therefore we propose a new benchmark, \textsc{LAMA-ckl}, to address this issue. Through experiments conducted on both newly introduced and established CKL benchmarks, TAALM proves the state-of-the-art performance upon the baselines, and also shows synergistic compatibility when integrated with previous CKL approaches.

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。

↑