New MCMC algorithm reduces subset selection passes to 2 for optimal k-dimensional subspace approximation.
problem Subset selection for k-dimensional subspace approximation with ε-approximation. method MCMC sampling algorithm reducing passes to 2 for p=2 case, poly(k/ε) size subset. result Subset selection of nearly optimal size in 2 passes, (1+ε) approximation. Improved DDPMs achieve high log-likelihoods and sample quality with fewer passes.
problem Improving log-likelihoods of DDPMs while maintaining high sample quality.
method Modifying DDPMs with learning variances and using precision-recall metrics.
result DDPMs can achieve competitive log-likelihoods with fewer forward passes.
New algorithms accelerate MAP inference in Markov fields with faster convergence.
problem Finding the most likely configuration in discrete-valued Markov random fields.
method Entropy-regularized linear programming with accelerated gradient methods.
result Accelerated algorithms find optimal solutions faster, especially when the LP is tight.
DimeNet uses directional message passing to improve molecular predictions.
problem Lack of directional information in graph neural networks for molecules.
method Directional message passing, rotationally equivariant embeddings, spherical functions.
result DimeNet outperforms previous GNNs by 76% on MD17 and 31% on QM9.
Introduces Spectral Graph Network combining spatial and spectral message passing.
problem Relational reasoning in graph structured data.
method Applies message passing to both spatial and spectral domains of a graph.
result Promotes efficient training with fewer iterations and robustness to edge dropout.
We reduce training time in convolutional networks (CNNs) with a method that, for some of the mini-batches: a) scales down the resolution of input images via downsampling, and b) reduces the forward pass operations via pooling on the convolution filters. Training is performed in an interleaved fashion; some batches unde…
LLoCa makes any network Lorentz-equivariant, achieving high accuracy and efficiency.
problem Limitations of specialized layers in Lorentz-equivariant neural networks.
method LLoCa framework using local reference frames and geometric message passing.
result Models achieve competitive and state-of-the-art accuracy on particle physics tasks.
Graph neural networks improve with affinity measures from random walks.
problem Limited expressivity of GNNs due to small receptive field.
method Introduced affinity measures from random walks into GNNs.
result Affinity measures enhance GNN performance on various tasks.
Noisy Pooled PCR tests large groups more efficiently.
problem Efficiently test large populations for viral infections.
method Converts group testing to a linear inverse problem with a message passing algorithm.
result Estimates patient illness status with fewer pooled measurements.
New method reduces diffusion model function evaluations for discrete data.
problem High computational burden in generating samples from masked diffusion models.
method Modified causal attention mask and speculative sampling mechanism for non-factorized predictions.
result Achieved ~2x reduction in required network forward passes.
Deep structured output learning shows great promise in tasks like semantic image segmentation. We proffer a new, efficient deep structured model learning scheme, in which we show how deep Convolutional Neural Networks (CNNs) can be used to estimate the messages in message passing inference for structured prediction wit…
MPNNs struggle with class-bottlenecks and heterophily, leading to performance limitations.
problem Performance limitations of MPNNs under heterophily and structural bottlenecks.
method A statistical framework decomposing model performance into SNR components and proving bounds on sensitivity.
result Optimal graph structures for maximizing higher-order homophily are disjoint unions of single-class and two-class-bipartite clusters.
New method improves image denoising with fewer parameters and less data.
problem Image denoising requires large datasets and supervised settings, limiting practical applications.
method Self-supervised framework using Tucker low-rank tensor approximation.
result Improves model generalizability and reduces data acquisition costs.
Study finds exact limits for sparse regression with fewer observations than usual.
problem Understanding sparse linear regression with sublinear sparsity.
method Adaptive interpolation method and modified AMP algorithm.
result Exact asymptotic expressions for mutual information and MMSE in sublinear sparsity.
We propose a novel approach for semantic segmentation that uses an encoder in the reverse direction to decode. Many semantic segmentation networks adopt a feedforward encoder-decoder architecture. Typically, an input is first downsampled by the encoder to extract high-level semantic features and continues to be fed for…
A new GNN module learns geometric scattering features for better graph classification and feature exploration.
problem Learning long-range graph relations and extracting meaningful features from graphs.
method Proposes a learnable geometric scattering (LEGS) module in graph neural networks (GNNs), incorporating wavelet filters.
result LEGS-based GNNs outperform existing methods in graph classification and feature extraction tasks.
Seminal works on graph neural networks have primarily targeted semi-supervised node classification problems with few observed labels and high-dimensional signals. With the development of graph networks, this setup has become a de facto benchmark for a significant body of research. Interestingly, several works have rece…
Predicting human behavior is a difficult and crucial task required for motion planning. It is challenging in large part due to the highly uncertain and multi-modal set of possible outcomes in real-world domains such as autonomous driving. Beyond single MAP trajectory prediction, obtaining an accurate probability distri…
Top-KAST maintains constant sparsity in large neural networks, improving performance and reducing resource usage.
problem Training large sparse neural networks is computationally expensive and resource-intensive.
method Top-KAST preserves constant sparsity in both forward and backward passes during training.
result Top-KAST outperforms previous methods on the ImageNet benchmark and language modeling tasks.
Multigrid modeling algorithms are a technique used to accelerate relaxation models running on a hierarchy of similar graphlike structures. We introduce and demonstrate a new method for training neural networks which uses multilevel methods. Using an objective function derived from a graph-distance metric, we perform or…
We propose a novel online learning algorithm for Restricted Boltzmann Machines (RBM), namely, the Online Generative Discriminative Restricted Boltzmann Machine (OGD-RBM), that provides the ability to build and adapt the network architecture of RBM according to the statistics of streaming data. The OGD-RBM is trained in…
Efficiently computes matrix square roots and their inverses for large matrices.
problem Computing matrix square roots and inverses for large matrices efficiently.
method Combines Krylov subspace methods with rational approximation for quadratic-time computation.
result Achieves 4 decimal places of accuracy with fewer than 100 matrix-vector multiplications.
We analyze oversquashing in topological message-passing using relational structures.
problem Oversquashing in topological message-passing remains understudied.
method A unifying axiomatic framework that bridges graph and topological message-passing.
result Potential to advance topological deep learning.
Message passing is the key to graph neural networks, but new terms are needed to avoid confusion.
problem Current methods of graph neural networks cannot solve all problems over given input graphs.
method Demonstrates that any function of interest can be expressed using pairwise message passing over a modified graph.
result Message passing is the fundamental approach for graph neural networks, and new terms are needed to avoid confusion.
Optimizes reinforcement learning by prioritizing sets of samples over individual ones.
problem Limits exploration and improvement on harder examples due to focusing on isolated samples.
method Proposes Pass@K Policy Optimization (PKPO) to optimize for sets of samples that maximize reward when considered jointly.
result Optimization with novel low variance unbiased estimators for pass@k and its gradient leads to significant pass@k gains.
The study predicts pass completion probability in NFL games.
problem Estimating the likelihood of NFL pass completion.
method Machine learning algorithm using distance measures and random forest model.
result Developed a method to predict pass completion probability.
This article is motivated by soccer positional passing networks collected across multiple games. We refer to these data as replicated spatial passing networks---to accurately model such data it is necessary to take into account the spatial positions of the passer and receiver for each passing event. This spatial regist…
ADMP-GNN dynamically adjusts message-passing layers for better graph learning performance.
problem Fixed message-passing steps in GNNs do not account for nodes' varying computational needs.
method Proposes ADMP-GNN, which dynamically adjusts the number of message-passing layers for each node.
result Improves performance on node classification tasks compared to baseline GNN models.
A new knot move preserves pass-move equivalence and differs in count.
problem Defining and analyzing a new knot move.
method Introducing the 1-2-move and comparing it to pass-move and #-move. result Equivalence under 1-2-move for knots is equivalent to pass-move equivalence. Every classical knot is band-pass equivalent to the unknot or the trefoil. The band-pass class of a knot is a concordance invariant. Every ribbon knot, for example, is band-pass equivalent to the unknot. Here we introduce the long virtual knot concordance group VC. It is shown that for every concordance cla…
Factor graphs have recently gained increasing attention as a unified framework for representing and constructing algorithms for signal processing, estimation, and control. One capability that does not seem to be well explored within the factor graph tool kit is the ability to handle deterministic nonlinear transformati…
The Turaev genus and dealternating number of a link are two invariants that measure how far away a link is from alternating. We determine the Turaev genus of a torus knot with five or fewer strands either exactly or up to an error of at most one. We also determine the dealternating number of a torus knot with five or f…
Complete classification of links up to specific moves.
problem Classifying links using specific local moves.
method Clasper theory.
result Complete classification of links up to clasp-pass moves.
We consider stochastic gradient descent (SGD) for least-squares regression with potentially several passes over the data. While several passes have been widely reported to perform practically better in terms of predictive performance on unseen data, the existing theoretical analysis of SGD suggests that a single pass i…
New graph coarsening method preserves GNN message-passing signals.
problem Preserving GNN message-passing on coarsened graphs.
method Proposed a new oriented message-passing operation for coarsened graphs.
result Improved node classification results compared to naive methods.
Paper tackles non-uniform coverage planning for robots.
problem Non-uniform coverage planning for robots that need to visit some points more frequently.
method Proposes a novel reinforcement learning approach in a Semi-Markov Decision Process.
result Significant improvement over existing greedy approach in simulations.
Study on convergence of graph neural networks on random graphs.
problem Convergence of message passing graph neural networks on large random graphs.
method Extended convergence results to a broad class of aggregation functions using McDiarmid inequality.
result Non-asymptotic bounds for convergence quantified with high probability.
Quantitative analysis of soccer players' passing ability focuses on descriptive statistics without considering the players' real contribution to the passing and ball possession strategy of their team. Which player is able to help the build-up of an attack, or to maintain the possession of the ball? We introduce a novel…
DG improves policy gradient efficiency by selectively backpropagating only valuable samples.
problem Expensive backward passes in policy gradient methods reduce efficiency.
method Introduces 'delight' as a forward-pass signal of learning value and a Kondo gate to selectively backpropagate.
result Selective backpropagation reduces backward pass costs without sacrificing learning quality.
Stochastic Gradient Descent can overfit after just a few passes, contrary to initial expectations.
problem Understanding the out-of-sample performance of multi-pass SGD in stochastic convex optimization.
method Analysis of multi-pass SGD in the stochastic convex optimization model.
result Multi-pass SGD can lead to significant overfitting after just a few passes, contrary to initial expectations.
One-pass algorithm finds small subset for ℓp subspace approximation with additive error.
problem Finding a small subset of data points for ℓp subspace approximation. method One-pass subset selection with additive approximation guarantee for p∈[1,∞). result First one-pass algorithm with additive error for ℓp subspace approximation. The (ordinary) unknotting-number of 1-dimensional knots, which is defined by using the crossing-change, is a very basic and important invariant. It is very natural to consider the `unknotting-number' associated with other local-moves on n-dimensional knots, where n is a natural number. In this paper we prove the follow…
Study virtualized Delta, Sharp, and Pass moves for oriented virtual knots and links.
problem Conditions for unknotting oriented virtual knots and links.
method Local moves (Delta, Sharp, Pass) for oriented virtual links.
result Lower bounds for the unknotting numbers are proven and shown to be best possible.
A guide to using low-pass graph filters for network data.
problem Understanding and processing graph data with low-pass filters.
method Definition and application of low-pass graph filters to graph data.
result Low-pass filters effectively retain lower frequency graph data contents.
We investigate minimal surfaces passing a given curve in R3. Using the Frenet frame of a given curve and isothermal parameter, we derive the necessary and sufficient condition for minimal surface. Also we derive the parametric representation of two minimal surface families passing a circle and a helix as examples.
New algorithm reduces memory usage for multi-pass bandit problems.
problem Memory-efficient multi-pass bandit algorithms for large action spaces.
method Develops a B-pass algorithm with O(1) memory that achieves optimal regret. result Sharp memory-regret trade-off: O(1) memory suffices for Θ(T1/2) regret in B passes. Full-batch GD outperforms one-pass SGD in learning a single-index model with quadratic activation.
problem Learning a single-index model with quadratic activation using gradient descent.
method Full-batch gradient descent compared to one-pass stochastic gradient descent (SGD) on a correlation loss.
result Full-batch GD requires only n≃d samples for strong recovery, while one-pass SGD requires n≳dlogd samples. We show that the proportion of hyperbolic knots among all of the prime knots of n or fewer crossings does not converge to 1 as n approaches infinity. Moreover, we show that if K is a nontrivial knot then the proportion of satellites of K among all of the prime knots of n or fewer crossings does not converge…