Kernel ridge regression (KRR) and Gaussian processes (GPs) are fundamental tools in statistics and machine learning, with recent applications to highly over-parameterized deep neural networks. The ability of these tools to learn a target function is directly related to the eigenvalues of their kernel sampled on the input data distribution. Targets that have support on higher eigenvalues are more learnable. However, solving such eigenvalue problems on real-world data remains a challenge. Here, we consider cross-dataset learnability and show that one may use eigenvalues and eigenfunctions associated with highly idealized data measures to reveal spectral bias on complex datasets and bound learnability on real-world data. This allows us to leverage various symmetries that realistic kernels manifest to unravel their spectral bias.
Demystifying Inductive Biases for $β$-VAE Based Architectures
Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所)
$β$ -变分自动编码器($β$ -VAE)及其变体在学习语义上有意义的解纠缠表示方面的性能是无与伦比的。另一方面,有理论论据表明,无监督的解缠是不可能的。在这项工作中,我们揭示了负责基于VAE的架构的成功的感性偏见。我们表明,在经典的数据集的方差结构,诱导的生成因子,是方便地对齐的潜在方向培育的VAE目标。这就建立了VAE的解开能力所依赖的关键偏差。通过对现有数据集进行小而精细的扰动,我们隐藏了方便的相关结构,这种结构很容易被各种架构所利用。为了证明这一点,我们构建了标准数据集的修改版本,其中(i)生成因子被完美地保留;(ii)每个图像都经过温和的变换,导致方差的微小变化;(iii)领先的\textbf{VAE-based disentanglement architectures fail to produce disentangled representations whilst the performance of a non-variational method remains unchanged}。我们的修改的建设是不平凡的,并依赖于最近的进展机制的理解$β$ -VAE及其连接到PCA。我们通过提供具有独立利益的额外见解来加强这种联系。
Department of Biostatistics, Columbia Mailman School of Public Health(哥伦比亚大学生物统计学系、哥伦比亚大学马利曼公共卫生学院);Institute for Data, Systems, and Society, Massachusetts Institute of Technology(麻省理工学院数据、系统和社会研究所);Department of Biostatistics, Yale School of Public Health(耶鲁大学生物统计学系)
Towards Understanding the Spectral Bias of Deep Learning
Department of Computer Science, University of California, Los Angeles(加利福尼亚大学洛杉矶分校计算机科学系);Department of Mathematics, City University of Hong Kong(香港城市大学数学系)
Overcoming the Spectral Bias of Neural Value Approximation
NSF AI Institute for Artificial Intelligence and Fundamental Interactions (IAIFI)(NSF人工智能与基本相互作用研究所(IAIFI));Computer Science and Artificial Intelligence Laboratory (CSAIL)(计算机科学与人工智能实验室(CSAIL));Improbable AI Lab(Improbable人工智能实验室);Massachusetts Institute Technology(麻省理工学院)