arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.05232cs.LGcs.AI
收藏

通用神经泛函

Universal Neural Functionals

Stanford University(斯坦福大学);Google DeepMind(谷歌DeepMind)

Allan Zhou, Chelsea Finn, James Harrison

AI总结:

本文提出通用神经泛函(UNFs),一种自动构建任意权重空间置换等变模型的算法,可替换进学习优化器,在图像分类和语言模型优化中优于先前方法。

摘要:

许多现代机器学习任务中的一个挑战性问题是处理权重空间特征,即从神经网络的权重和梯度中转换或提取信息。近期研究开发了有前景的权重空间模型,这些模型对简单前馈网络的置换对称性具有等变性。然而,这些模型不适用于一般架构,因为权重空间的置换对称性可能因循环连接或残差连接而变得复杂。本文提出了一种算法,能够为任意权重空间自动构建置换等变模型,我们称之为通用神经泛函(UNFs)。在其他应用中,我们展示了如何将UNFs替换到现有的学习优化器设计中,并在优化小型图像分类器和语言模型时发现相较于先前方法有显著改进。我们的结果表明,学习优化器可以受益于考虑其优化的权重空间的(对称)结构。我们在https://github.com/AllanYangZhou/universal_neural_functional开源了构建UNFs的库。

英文摘要:

A challenging problem in many modern machine learning tasks is to process weight-space features, i.e., to transform or extract information from the weights and gradients of a neural network. Recent works have developed promising weight-space models that are equivariant to the permutation symmetries of simple feedforward networks. However, they are not applicable to general architectures, since the permutation symmetries of a weight space can be complicated by recurrence or residual connections. This work proposes an algorithm that automatically constructs permutation equivariant models, which we refer to as universal neural functionals (UNFs), for any weight space. Among other applications, we demonstrate how UNFs can be substituted into existing learned optimizer designs, and find promising improvements over prior methods when optimizing small image classifiers and language models. Our results suggest that learned optimizers can benefit from considering the (symmetry) structure of the weight space they optimize. We open-source our library for constructing UNFs at https://github.com/AllanYangZhou/universal_neural_functional.

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。

↑