Transformers learn low-dimensional target functions efficiently in-context.
problem Efficiently learning nonlinear target functions in-context using transformers.
method Nonlinear MLP layer in transformers optimized by gradient descent, focusing on single-index target functions.
result Transformers can learn target functions with low-dimensional structures efficiently in-context.
The paper shows diffusion models can converge faster to a target distribution with low-dimensional structure.
problem Improving the convergence rate of diffusion models to target distributions.
method Analyzing DDIM and DDPM samplers under low-dimensional structure assumptions.
result The iteration complexities of DDIM and DDPM are no greater than k/ε in total variation distance. Deep neural networks can estimate Q-values efficiently on low-dimensional state-action spaces.
problem Estimating the performance of a reinforcement learning policy using data from a different policy.
method Deep fitted Q-evaluation method leveraging manifold structure and convolutional neural networks.
result Sharp error bound for fitted Q-evaluation depends on intrinsic dimension and function space mismatch.
Localized diffusion models reduce training complexity by exploiting low-dimensional structure.
problem Training diffusion models is computationally expensive due to the curse of dimensionality.
method Localized neural networks and localized score matching loss to estimate low-dimensional score functions.
result Localized diffusion models can circumvent the curse of dimensionality with reduced sample complexity.
Paper adapts DDPM to low-dimensional structures in image distributions.
problem Understanding and adapting to low-dimensional structures in image distributions.
method Developed a novel set of analysis tools to characterize algorithmic dynamics.
result First theoretical demonstration that DDPM can adapt to unknown low-dimensional structures.
Mamba efficiently learns low-dimensional targets in-context via feature extraction.
problem Learning low-dimensional targets in context for computational efficiency.
method Test-time feature learning of a single-index model using Mamba's pretrained linear-time sequence model.
result Mamba achieves efficient in-context learning of low-dimensional targets via feature extraction.
LASE improves local network structure visualization by targeting locally low-dimensional regions.
problem Global spectral embedding fails to capture local geometric features in sparse, transitive networks.
method Local Adjacency Spectral Embedding (LASE) using weighted spectral decomposition.
result LASE reveals locally low-dimensional structure, improving local reconstruction and visualization.
Paper explores how Rectified Flow adapts to low-dimensional data.
problem Improving sampling efficiency in low-dimensional data.
method Investigates Rectified Flow's adaptation to low-dimensional support and introduces a stochastic version.
result Shows improved sampling efficiency with O(k/ε) complexity. Non-linear dimensionality reduction techniques such as manifold learning algorithms have become a common way for processing and analyzing high-dimensional patterns that often have attached a target that corresponds to the value of an unknown function. Their application to new points consists in two steps: first, embedd…
A scalable method for accurate inference of low-dimensional parameters in high-dimensional linear regression.
problem Statistical inference for low-dimensional parameters in high-dimensional linear regression models.
method Mean-field variational Bayes approach, focusing on nuisance parameters and conditional distributions.
result Competitive numerical performance and theoretical guarantees for estimation and uncertainty quantification.
It is known that any target function is realized in a sufficiently small neighborhood of any randomly connected deep network, provided the width (the number of neurons in a layer) is sufficiently large. There are sophisticated theories and discussions concerning this striking fact, but rigorous theories are very compli…
Study on reliability of latent reuse in diffusion models under distribution shift.
problem When can latent spaces from a source dataset be reused for a target dataset with different distributions?
method Considered a source-target setting with approximately low-dimensional datasets near different subspaces. Analyzed the target-domain score error due to principal-angle misalignment and target ambient noise.
result Latent reuse is reliable only if the source and target subspaces are close and the target ambient noise is not too amplified.
Data repetition improves SGD's learning of high-dimensional functions.
problem Learning pertinent features in multi-index models with high-dimensional noisy data.
method Investigation of two-layer shallow neural networks trained with gradient-based algorithms, focusing on data repetition.
result Data repetition significantly improves the computational efficiency of SGD, learning all directions with at most O(dlogd) steps. Adversarial robustness in multi-index models is as easy as standard learning.
problem Adversarial robustness in high-dimensional multi-index models.
method Proves that hidden directions of multi-index models offer a Bayes optimal low-dimensional projection for robustness against ℓ2-bounded adversarial perturbations. result Adversarially robust learning is as easy as standard learning, requiring no additional samples.
Analyzes SGD dynamics in two-layer networks, bridging different regimes.
problem Understanding SGD dynamics in high-dimensional and mean-field settings.
method Rigorous analysis via deterministic low-dimensional description of sufficient statistics.
result Infinite-width dynamics remains close to a low-dimensional subspace.
New method uses limited labeled data and multiple starts to adapt models across domains.
problem Accurate predictions in target domain with few labeled data.
method Fine-tuning from multiple adaptive starts, extending UDA methods.
result Minimax-optimal target performance with limited labeled target data.
The increasing use of machine-learning (ML) enabled systems in critical tasks fuels the quest for novel verification and validation techniques yet grounded in accepted system assurance principles. In traditional system development, model-based techniques have been widely adopted, where the central premise is that abstr…
Deep networks can approximate high-dimensional distributions from low-dimensional ones.
problem Approximating high-dimensional distributions from low-dimensional ones.
method Proved neural networks can transform low-dimensional distributions to high-dimensional ones with arbitrary closeness measured by Wasserstein distances and maximum mean discrepancy.
result Upper bounds of the approximation error are obtained in terms of the width and depth of neural network.
SGD trains NNs to learn low-dimensional representations efficiently.
problem Training two-layer NNs with SGD for multiple-index models.
method Online SGD with weight decay for training.
result First-layer weights converge to the principal subspace of the true model.
Graph-Laplacians and their spectral embeddings play an important role in multiple areas of machine learning. This paper is focused on graph-Laplacian dimension reduction for the spectral clustering of data as a primary application. Spectral embedding provides a low-dimensional parametrization of the data manifold which…
The paper improves the probability flow ODE sampler for faster sampling of natural images.
problem Improving the convergence rate of the probability flow ODE sampler.
method Adapting the probability flow ODE sampler to exploit intrinsic low-dimensional structures in natural image data.
result Achieves a dimension-free convergence rate of O(k/T) in total variation distance, improving upon existing results. The paper analyzes reflected diffusion models on hypercube data.
problem Challenges in modeling bounded domains with low-dimensional data.
method Employed an infinite series expansion of transition densities to bound the score function and its approximation.
result Established convergence rates for generative algorithm adapting to intrinsic dimensionality.
Paper tackles robust domain adaptation without target domain data.
problem Learning domain invariant representations without target domain data.
method Integrates deep autoencoder and causal structure learning into a unified model.
result CAE learns causal representations using only source domain data.
Gradient flow solves multi-index regression for high-dimensional Gaussian data.
problem Learning multi-index functions from high-dimensional Gaussian data.
method Two-timescale algorithm with non-parametric link function learning.
result Global convergence of Grassmannian population gradient flow dynamics.
Two-layer networks learn faster with batch reuse, overcoming information and leap exponents.
problem Limitations of gradient flow and single-pass GD in learning multi-index target functions.
method Multi-pass gradient descent that reuses batches, analyzed using Dynamical Mean-Field Theory.
result Two-time-step overlap with target subspace for non-staircase functions, overcoming information and leap exponents.
We investigate the low-dimensional structure of deterministic transformations between random variables, i.e., transport maps between probability measures. In the context of statistics and machine learning, these transformations can be used to couple a tractable "reference" measure (e.g., a standard Gaussian) with a tar…
Neural network learns low-dimensional polynomials with SGD near information-theoretic limit.
problem Learning a single-index target function with gradient descent.
method Two-layer neural network optimized by SGD on squared loss.
result Sample and runtime complexity of n≃T=Θ(d⋅polylogd) for polynomial single-index models, matching information theoretic limit up to polylogarithmic factors. The paper tackles high-dimensional Bayesian optimization using tree-structured additive models.
problem Scaling Bayesian Optimization to high-dimensional problems.
method Tree-structured additive models with hybrid graph learning and zooming-based algorithms.
result Demonstrates faster model learning and reduced model complexity in high-dimensional settings.
Explains how knots relate to 4D shapes.
problem Understanding 4D shapes through knot theory.
method Combines knot theory with 4D manifold topology.
result Connects 4D shapes to knot theory and other geometries.
Deep learning has exhibited superior performance for various tasks, especially for high-dimensional datasets, such as images. To understand this property, we investigate the approximation and estimation ability of deep learning on anisotropic Besov spaces. The anisotropic Besov space is characterized by direction-depen…
A new method for SVGD reduces variance in high dimensions.
problem High-dimensional variance in SVGD.
method Grassmann Stein Variational Gradient Descent (GSVGD) projects onto arbitrary subspaces and uses coupled Grassmann-valued diffusion.
result GSVGD explores high-dimensional problems with intrinsic low-dimensional structure efficiently.
Study how data structure impacts classification performance in overparametrized models.
problem Understanding the role of data structure in overparametrized models.
method Analytical and statistical physics methods applied to structured data.
result Data structure significantly affects the performance of overparametrized models, particularly for logistic loss.
Improved likelihood-free inference by localizing and refining low-dimensional approximations.
problem Poor performance of common likelihood-free methods in high-dimensional models.
method Localisation followed by refinement of low-dimensional summaries.
result Improved accuracy in marginal posteriors through localized and refined approximations.
Optimal batch size minimizes training time for neural networks.
problem Minimizing training time for two-layer neural networks with SGD.
method Characterized optimal batch size as a function of target hardness (information exponents). Used Correlation loss SGD to overcome limitations.
result Optimal batch size minimizes training time without changing total sample complexity.
Proves Kastler-Kalau-Walze theorem for spectral Einstein functional on low-dimensional manifolds.
problem Proving Kastler-Kalau-Walze type theorems for spectral Einstein functional.
method Defining spectral Einstein functional associated with Dirac operator and proving theorem for low-dimensional manifolds.
result Proves Kastler-Kalau-Walze type theorem for spectral Einstein functional on low-dimensional manifolds with boundary.
This work refines Cover's theory for binary classification on low-dimensional data.
problem The challenge of analyzing how low-dimensional data structures affect classification models.
method Refines Cover's function-counting theory to account for low-dimensional data structure.
result Derives dichotomy counts and analyzes the impact of data structure on classification models.
Paper shows robust generative learning with minimal assumptions on target distributions.
problem Learning generative models with minimal assumptions on target distributions.
method Lipschitz-regularized α-divergences with minimal assumptions. result Stable learning across various target distributions with minimal assumptions.
Deep reinforcement learning achieves superhuman performance in a range of video game environments, but requires that a designer manually specify a reward function. It is often easier to provide demonstrations of a target behavior than to design a reward function describing that behavior. Inverse reinforcement learning …
The paper provides convergence guarantees for ODE-based generative models using transformers.
problem Theoretical guarantees for ODE-based generative models.
method A pre-trained autoencoder maps inputs to a latent space, and a transformer predicts the velocity field.
result The distribution of samples generated via estimated ODE flow converges to the target distribution in Wasserstein-2 distance.
New neural network with RePU activation approximates smooth functions and their derivatives.
problem Approximating smooth functions and their derivatives with neural networks.
method Differentiable neural networks with RePU activation functions.
result Improved approximation error bounds for RePU-activated neural networks.
Wide neural networks can outperform kernel methods in certain tasks.
problem Understanding when neural networks outperform kernel methods in classification tasks.
method Analyzing the performance of wide neural networks and kernel methods on various tasks, considering the initialization of SGD and the structure of covariates.
result Wide neural networks can outperform kernel methods in tasks where covariates have a low-dimensional structure similar to the target function.
Study on infinitely-wide CNNs and their adaptability to function spatial scales.
problem Understanding how CNNs efficiently learn high-dimensional functions and their adaptability to function spatial scales.
method Study infinitely-wide deep CNNs in the kernel regime, characterizing their spectrum and using generalisation bounds to prove adaptability.
result Deep CNNs adapt to the spatial scale of the target function, with error decay controlled by the effective dimensionality of function subsets.
Method identifies low-dimensional structure in high-dimensional probability measures.
problem Identifying low-dimensional structure in high-dimensional probability measures.
method Extends prior work on minimizing majorizations of the Kullback-Leibler divergence to identify optimal approximations within a specific class of measures.
result Connection between dimensional logarithmic Sobolev inequality and approximations with the ansatz.
Real world data often exhibit low-dimensional geometric structures, and can be viewed as samples near a low-dimensional manifold. This paper studies nonparametric regression of Hölder functions on low-dimensional manifolds using deep ReLU networks. Suppose n training data are sampled from a Hölder function in $\mathc…
Shallow diffusion models learn hidden low-dimensional structures effectively.
problem Learning from high-dimensional signals like images and video.
method Analysis of shallow diffusion models over the Barron space of single layer neural networks.
result Shallow diffusion models can adapt to simple low-dimensional structures, overcoming the curse of dimensionality.
Deep, wide ConvResNets can approximate functions and their smoothness.
problem Function approximation and smoothness in deep networks.
method Analyzing ConvResNets, proving their ability to approximate functions and their smoothness.
result Large ConvResNets can approximate functions and exhibit sufficient first-order smoothness.
KALE flow approximates KL divergence for distributions with disjoint support.
problem Approximating KL divergence for distributions with disjoint support.
method Relaxed KL gradient flow using RKHS, continuously interpolating between KL and MMD.
result Global convergence of KALE flow under sufficient smoothness assumptions.
The paper tackles transfer learning for growing matrix representations, improving estimation accuracy.
problem Structured matrix estimation under growing ambient dimensions and latent representations.
method Proposes a general transfer framework decomposing target parameters into embedded source components, low-rank innovations, and sparse edits. Develops an anchored alternating projection estimator.
result Establishes deterministic error bounds that separate target noise, representation growth, and source estimation error, yielding improved rates.