Model trains passing events on a bridge using multilevel Gaussian process.
problem Represent aggregate train-passing events from a bridge monitoring system.
method Formulate a combined model with low-rank approximation hierarchical Gaussian process, incorporating domain expertise as constraints.
result Allow for simulation of previously unobserved train types.
A training-free message passing module improves hypergraph neural networks.
problem Computational intensive training process in hypergraph neural networks.
method Decoupling hypergraph structural information from model learning stage.
result TF-HNN achieves superior performance and training efficiency.
Unified framework for training SNNs using EP, faster convergence.
problem Training spiking neural networks with Expectation-Propagation.
method Message-passing framework for learning marginal distributions of SNN parameters.
result Faster convergence compared to gradient-based methods.
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.
One-pass optimisation for high-dimensional hyperparameters.
problem Efficient optimisation of hyperparameters in machine learning models.
method Approximate hypergradient-based optimisation for any continuous hyperparameter, requiring only one training episode.
result Competitive performance on various datasets without hyperparameter restarts.
SPACR trains uncertainty-aware regressors directly within a single pass, improving efficiency and validity.
problem Training uncertainty-aware regressors while maintaining efficiency and validity.
method Joint optimization of efficiency and validity during training.
result SPACR consistently provides tighter intervals and better coverage-efficiency trade-offs compared to standard CP and DOICR.
DOPPLER optimizes DP training with low-pass filtering, improving model accuracy.
problem Privacy concerns in deep learning models and performance degradation of DP optimizers.
method Developed DOPPLER, a low-pass filter for DP optimizers, to reduce privacy noise and enhance model quality.
result DOPPLER optimizers outperform non-DOPPLER counterparts by 3%-10% in test accuracy.
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…
Neural network training is computationally and memory intensive. Sparse training can reduce the burden on emerging hardware platforms designed to accelerate sparse computations, but it can affect network convergence. In this work, we propose a novel CNN training algorithm Sparse Weight Activation Training (SWAT). SWAT …
PASS protects private attributes by stochastically substituting data.
problem Protecting private attributes in ML services while maintaining data utility.
method PASS uses stochastic data substitution with a novel loss function derived from information theory.
result PASS effectively protects private attributes across various datasets.
ORFit trains models on streaming data with one pass, minimizing memory and computational costs.
problem Training large models on a stream of data without retraining on previous data.
method Orthogonal Recursive Fitting (ORFit) using orthogonal gradient descent and recursive least-squares.
result ORFit updates parameters orthogonally to past gradients, leading to efficient memory and computational usage.
This paper introduces Selective-Backprop, a technique that accelerates the training of deep neural networks (DNNs) by prioritizing examples with high loss at each iteration. Selective-Backprop uses the output of a training example's forward pass to decide whether to use that example to compute gradients and update para…
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.
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.
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.
DD algorithm tracks test error from train error without validation data.
problem Systematic generalization gap between train and test errors in modern model training.
method Decoupled descent (DD) algorithm that cancels data reuse biases via approximate message passing.
result DD algorithm rigorously demonstrates zero-cost validation and 100% data utilization.
Backpropagation-free RL method trains layers using local signals.
problem Vanishing or exploding gradients in backpropagation-based RL.
method Local pairwise distance matching for layer-wise training without backpropagation.
result Backpropagation-free method achieves competitive performance and stability.
Neural networks offer high-accuracy solutions to a range of problems, but are costly to run in production systems because of computational and memory requirements during a forward pass. Given a trained network, we propose a techique called Deep Learning Approximation to build a faster network in a tiny fraction of the …
Single-pass method estimates neural network uncertainty.
problem Uncertainty estimation in deep learning requires multiple passes.
method Probabilistic reasoning over neural network depths.
result Single forward pass for uncertainty estimation.
We introduce a new method for training deep Boltzmann machines jointly. Prior methods require an initial learning pass that trains the deep Boltzmann machine greedily, one layer at a time, or do not perform well on classifi- cation tasks.
SHINE uses forward pass quasi-Newton matrices to approximate Jacobian inverses for faster bi-level optimization.
problem Efficiently solving bi-level optimization problems with large Jacobian matrices.
method Proposes using quasi-Newton matrices from the forward pass to approximate the inverse Jacobian matrix.
result Empirically shows SHINE reduces computational cost of the backward pass for various problems.
Adapts BP-based algorithms for deep learning, improving performance and accuracy.
problem Training deep neural networks with discrete weights and activations.
method Message-passing algorithms based on Belief Propagation, with reinforcement field.
result Comparable performance to SGD-inspired heuristics (BinaryNet) and higher accuracy in predictions.
Improved scaling laws in linear regression using data reuse.
problem Sustainability of neural scaling laws when running out of new data.
method Data reuse in multi-pass stochastic gradient descent (multi-pass SGD) for M-dimensional linear models trained on N data with sketched features. result Multi-pass SGD achieves a test error of Θ(M1−b+L(1−b)/a) with L>N, improving scaling laws in data-constrained regimes. 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. Knowledge graph completion aims to predict missing relations between entities in a knowledge graph. In this work, we propose a relational message passing method for knowledge graph completion. Different from existing embedding-based methods, relational message passing only considers edge features (i.e., relation types)…
We propose a Bayesian nonparametric method for low-pass filtering that can naturally handle unevenly-sampled and noise-corrupted observations. The proposed model is constructed as a latent-factor model for time series, where the latent factors are Gaussian processes with non-overlapping spectra. With this construction,…
SNAPO optimizes policies for complex sequential decisions using differentiable simulation.
problem Optimizing policies for high-dimensional, sequential decisions under uncertainty.
method Embeds neural policy in a differentiable simulator, computes gradients efficiently.
result Produces sensitivities at a cost proportional to one reverse pass, regardless of sensitivity count.
A new neural network model for molecular graphs that learns efficiently and accurately.
problem Learning on molecular graphs with cycles and complex structures.
method Hierarchical inter-message passing using raw graph and junction tree representations.
result The model outperforms classical GNNs in detecting cycles and is efficient to train.
A fundamental computation for statistical inference and accurate decision-making is to compute the marginal probabilities or most probable states of task-relevant variables. Probabilistic graphical models can efficiently represent the structure of such complex data, but performing these inferences is generally difficul…
We propose Sideways, an approximate backpropagation scheme for training video models. In standard backpropagation, the gradients and activations at every computation step through the model are temporally synchronized. The forward activations need to be stored until the backward pass is executed, preventing inter-layer …
Exploiting sparsity enables hardware systems to run neural networks faster and more energy-efficiently. However, most prior sparsity-centric optimization techniques only accelerate the forward pass of neural networks and usually require an even longer training process with iterative pruning and retraining. We observe t…
One-pass SGD converges in overparametrized neural networks with random data.
problem Understanding convergence of SGD in neural networks with streaming data.
method Overparameterized two-layer neural networks, one-pass SGD, random initialization, NTK eigen-decomposition, VC dimension, McDiarmid's inequality.
result Prediction error converges in expectation under one-pass SGD in overparametrized neural networks.
New method estimates data influence efficiently by leveraging test samples.
problem Efficiently estimating influence of training data on model predictions.
method Mirrored Influence Hypothesis, forward pass for test samples.
result Significant improvement in efficiency over existing methods.
Training convolutional neural network models is memory intensive since back-propagation requires storing activations of all intermediate layers. This presents a practical concern when seeking to deploy very deep architectures in production, especially when models need to be frequently re-trained on updated datasets. In…
Approximate Message Passing (AMP) has been shown to be an excellent statistical approach to signal inference and compressed sensing problem. The AMP framework provides modularity in the choice of signal prior; here we propose a hierarchical form of the Gauss-Bernouilli prior which utilizes a Restricted Boltzmann Machin…
Bayesian method improves online NARMAX model identification.
problem Online identification of nonlinear systems with small sample sizes and low noise.
method Variational Bayesian inference using message passing algorithm for polynomial NARMAX models.
result Variational Bayesian estimator outperforms recursive and offline least-squares methods.
Survey of methods to train deep architectures without E2EBP.
problem Training deep architectures without end-to-end backpropagation.
method Modular training, weakly modular hybrids.
result Provably optimal alternatives to E2EBP can match or surpass E2EBP performance.
There is great interest in "saliency methods" (also called "attribution methods"), which give "explanations" for a deep net's decision, by assigning a "score" to each feature/pixel in the input. Their design usually involves credit-assignment via the gradient of the output with respect to input. Recently Adebayo et al.…
New algorithms improve binary neural network configurations.
problem Training binary neural networks efficiently.
method Stochastic message passing algorithms (BP and SP) for discrete inference.
result Stochastic BP and SP find better BNN configurations.
Training neural networks with verifiable robustness guarantees is challenging. Several existing approaches utilize linear relaxation based neural network output bounds under perturbation, but they can slow down training by a factor of hundreds depending on the underlying network architectures. Meanwhile, interval bound…
SGD converges with perturbed forward-backward passes, explained by geometric amplification.
problem Analyzing convergence of SGD with perturbed forward-backward passes in composite optimization.
method Characterized propagation and amplification of perturbations, derived convergence guarantees for non-convex and PL objectives.
result Perturbations cascade through the computational graph, affecting convergence order under specific conditions.
Novel framework uses synthetic data to quantify uncertainty in complex data.
problem Uncertainty quantification in complex, unstructured data.
method Perturbation-Assisted Sample Synthesis (PASS) and Perturbation-Assisted Inference (PAI) framework.
result Statistically guaranteed validity in inference, enhancing reliability of synthetic data.
LayerNorm transformers have dead directions that can be read from their parameters alone.
problem Locating dead directions in LayerNorm transformers
method Using the inverse-scale direction of LayerNorm affine parameters
result Predicted dead direction matches measured bottom singular direction
UCPO improves diversity in reinforcement learning models, maintaining high accuracy.
problem RLVR objectives often lead to diversity collapse, reducing coverage of correct solutions.
method UCPO adds a conditional uniformity penalty to GRPO, redistributing probability mass.
result UCPO improves Pass@K and diversity while maintaining competitive Pass@1 accuracy.
New method calibrates uncertainty in molecular property predictions.
problem Uncalibrated uncertainty estimates in molecular property predictions.
method Message Passing Neural Networks with calibrated probabilistic predictive distribution.
result Accurate molecular formation energy predictions with well-calibrated uncertainty.
Proposes PSCs for UQ in deep nets without retraining.
problem Estimating uncertainty in deep nets with a single pass.
method Identifies sensitive, smooth intermediate layer, fits probabilistic model.
result PSCs achieve UQ and OOD detection performance matching existing methods.
Study shows gMPNNs struggle with OOD link prediction in larger test graphs.
problem Inductive out-of-distribution link prediction in larger test graphs.
method Theoretical analysis and development of a gMPNN with structural pairwise embeddings.
result Structural node embeddings from gMPNNs converge to random guessing as test graphs grow.
Proposes an alternative method to train RBMs with binary synapses using Bayesian learning rule.
problem Training RBMs with binary synapses is challenging due to discrete nature of synapses.
method Proposes an alternative optimization method using the Bayesian learning rule, updating natural parameters instead of expectation parameters.
result No additional clipping is needed as natural parameters take values in the entire real domain.