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 2602.04233stat.MLcs.LG
收藏

预训练模型规模扩大时目标采样复杂度的可证改进

Provable Target Sample Complexity Improvements as Pre-Trained Models Scale

University of Tsukuba(筑波大学);RIKEN AIP(理化学研究所AIP);SB Intuitions Corp.(SB Intuitions公司);Kyoto University(京都大学);Shiga University(滋贺大学);Institute of Science Tokyo(东京科学大学)

Kazuto Fukuchi, Ryuichiro Hataya, Kota Matsui

AI总结:

本文通过引入caulking框架,证明预训练模型规模扩大可降低下游任务的样本复杂度,为扩展定律提供了理论支持。

摘要:

预训练模型已成为在广泛下游任务中高效构建模型不可或缺的工具。通过经验研究揭示了扩展定律的优势,表明更大的预训练模型可以显著减少下游学习的样本复杂度。然而,现有对预训练模型的理论研究缺乏解释这一现象的能力。在本文中,我们通过引入一个名为caulking的新框架进行理论研究,该框架受到参数高效微调(PEFT)方法如基于适配器的微调、低秩适应和部分微调的启发。我们的分析证明了改进的预训练模型确实能够降低下游任务的样本复杂度,从而为经验观察到的预训练模型大小与下游性能之间的扩展关系提供了理论依据,而这一关系在现有结果中并未被涵盖。

英文摘要:

Pre-trained models have become indispensable for efficiently building models across a broad spectrum of downstream tasks. The advantages of pre-trained models have been highlighted by empirical studies on scaling laws, which demonstrate that larger pre-trained models can significantly reduce the sample complexity of downstream learning. However, existing theoretical investigations of pre-trained models lack the capability to explain this phenomenon. In this paper, we provide a theoretical investigation by introducing a novel framework, caulking, inspired by parameter-efficient fine-tuning (PEFT) methods such as adapter-based fine-tuning, low-rank adaptation, and partial fine-tuning. Our analysis establishes that improved pre-trained models provably decrease the sample complexity of downstream tasks, thereby offering theoretical justification for the empirically observed scaling laws relating pre-trained model size to downstream performance, a relationship not covered by existing results.

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。

↑