Optimizes pruning masks for neural networks using probabilistic fine-tuning and PAC-Bayes bounds.
problem Improving neural network performance through adaptive pruning of weights.
method Optimizes stochastic pruning masks by minimizing expected loss, considering data-adaptive regularization and feature alignment.
result Probabilistic fine-tuning leads to improved test error over baseline methods in neural networks.
XTNet estimates complex cross-treatment effects in multi-category, multi-valued settings.
problem Challenges in estimating causal effects for multi-category, multi-valued treatments.
method Dynamic Neural Masking for capturing treatment interactions without restrictive assumptions.
result XTNet consistently outperforms state-of-the-art baselines in multi-category, multi-valued treatment effect estimation.
The ConditionaL Neural Networks (CLNN) and the Masked ConditionaL Neural Networks (MCLNN) exploit the nature of multi-dimensional temporal signals. The CLNN captures the conditional temporal influence between the frames in a window and the mask in the MCLNN enforces a systematic sparseness that follows a filterbank-lik…
NeuroPaint infers missing brain area dynamics from multi-animal datasets.
problem Leveraging multi-animal datasets to understand interactions between brain areas.
method Masked autoencoding approach trained across animals with partial observations.
result Models can successfully reconstruct dynamics of unrecorded brain areas.
Automatic feature extraction using neural networks has accomplished remarkable success for images, but for sound recognition, these models are usually modified to fit the nature of the multi-dimensional temporal representation of the audio signal in spectrograms. This may not efficiently harness the time-frequency repr…
Study different masking schemes for a universal marginaliser.
problem Understand how well a neural approximator learns conditional distributions.
method Compare networks trained with various masking schemes.
result Neural approximators perform differently based on the masking scheme.
New study on guidance in masked diffusion models, showing how it shapes sampling dynamics.
problem Understanding how guidance influences the sampling behavior of masked diffusion models.
method Derived explicit solution to guided reverse dynamics, analyzing effects in 1D and 2D.
result Guidance amplifies class-specific regions and suppresses shared regions, affecting covariance structures.
We present the ConditionaL Neural Network (CLNN) and the Masked ConditionaL Neural Network (MCLNN) designed for temporal signal recognition. The CLNN takes into consideration the temporal nature of the sound signal and the MCLNN extends upon the CLNN through a binary mask to preserve the spatial locality of the feature…
Neural network based architectures used for sound recognition are usually adapted from other application domains such as image recognition, which may not harness the time-frequency representation of a signal. The ConditionaL Neural Networks (CLNN) and its extension the Masked ConditionaL Neural Networks (MCLNN) are des…
Neural network interpretability is a vital component for applications across a wide variety of domains. In such cases it is often useful to analyze a network which has already been trained for its specific purpose. In this work, we develop a method to produce explanation masks for pre-trained networks. The mask localiz…
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.
The ConditionaL Neural Network (CLNN) exploits the nature of the temporal sequencing of the sound signal represented in a spectrogram, and its variant the Masked ConditionaL Neural Network (MCLNN) induces the network to learn in frequency bands by embedding a filterbank-like sparseness over the network's links using a …
Dropout and RaM become equivalent in large ResNets as depth and width increase.
problem Improving performance in deep learning models.
method Comparing Dropout and Random Gradient Masking in ResNets.
result Dropout and RaM converge to the same large-scale limiting dynamics in ResNets.
Node Masking improves GNNs' scalability and generalization.
problem Improving GNNs' ability to handle arbitrary graphs.
method Introducing Node Masking to enhance GNNs' performance.
result Node Masking enables GNNs to generalize and scale better.
We present an unsupervised training approach for a neural network-based mask estimator in an acoustic beamforming application. The network is trained to maximize a likelihood criterion derived from a spatial mixture model of the observations. It is trained from scratch without requiring any parallel data consisting of …
LOCAL learns dynamic causal structures from time series data efficiently.
problem Challenges in discovering DAG from time series data due to dynamic nature and nonlinear interactions.
method LOCAL proposes a quasi-maximum likelihood-based score function and adaptive modules ACML and DGPL.
result LOCAL significantly outperforms existing methods in dynamic causal discovery.
Dynamic Sparse Training finds efficient sparse networks from scratch.
problem Finding efficient sparse neural networks.
method Jointly optimizes network parameters and sparsity with trainable thresholds.
result Achieves state-of-the-art performance with minimal performance loss.
Deep neural network architectures designed for application domains other than sound, especially image recognition, may not optimally harness the time-frequency representation when adapted to the sound recognition problem. In this work, we explore the ConditionaL Neural Network (CLNN) and the Masked ConditionaL Neural N…
MARS automatically selects tensor decomposition ranks, improving performance in neural network tasks.
problem Determining optimal decomposition ranks in tensor decompositions.
method MARS uses binary masks to learn optimal tensor structure during training via relaxed MAP estimation.
result MARS achieves better results than previous methods in various tasks.
A technique scales symbolic methods with gradients for neural model explanation.
problem Limited scalability of symbolic methods for large neural networks.
method Combines gradient-based methods with symbolic techniques using Integrated Gradients to focus on a subset of neurons.
result Produces sparser and higher saliency regions compared to gradient-based methods alone.
New analysis shows how attention masks and LayerNorm prevent rank collapse in transformers.
problem Rank collapse in transformer models with increasing depth.
method General analysis of rank collapse under self-attention, considering attention masks and LayerNorm.
result Self-attention with LayerNorm can prevent rank collapse and maintain a rich set of equilibria.
Differentiable Masking reveals how neural models make decisions across layers.
problem Intractable and expensive approximate search for input relevance in deep models.
method Differentiable Masking learns to mask inputs while maintaining differentiability.
result Reveals how decisions are formed across network layers in BERT models.
Paper proposes a new sparse Bayesian neural network for simpler, more efficient DNNs.
problem Complex and large DNN architectures require simplification for better performance and efficiency.
method Masked Bayesian Neural Networks (BNN) with nodewise sparsity and optimal posterior distributions.
result The proposed BNN yields well-condensed DNN architectures with similar accuracy and uncertainty quantification to large DNNs.
Structured State-Space Duality connects SSMs to masked attention.
problem Connecting SSMs and attention mechanisms for efficient modeling.
method Formalizing and generalizing SSD from scalar-identity to diagonal state matrices.
result Diagonal SSMs match training complexity lower bounds and support richer dynamics.
New method for efficient probabilistic inference using masked language modeling.
problem Efficient posterior inference in probabilistic programs with many hyper-parameters.
method Formulate inference as masked language modeling, train a neural network to unmask random values.
result Foundation posterior for zero-shot inference and fine-tuning across a range of programs.
AFS-BM improves model accuracy by dynamically selecting features.
problem Feature selection challenges in ML, especially scalability and adaptability.
method Joint optimization for feature selection and model training with binary masking.
result AFS-BM achieves significant improvements in model accuracy and computational efficiency.
We apply a fast kernel method for mask-based single-channel speech enhancement. Specifically, our method solves a kernel regression problem associated to a non-smooth kernel function (exponential power kernel) with a highly efficient iterative method (EigenPro). Due to the simplicity of this method, its hyper-parameter…
Unified model combines feature and label propagation for semi-supervised classification.
problem Combining feature and label propagation for effective semi-supervised classification.
method Unified Message Passing Model (UniMP) using Graph Transformer and masked label prediction.
result Obtains new state-of-the-art results in Open Graph Benchmark (OGB).
In this paper, we present a novel system that separates the voice of a target speaker from multi-speaker signals, by making use of a reference signal from the target speaker. We achieve this by training two separate neural networks: (1) A speaker recognition network that produces speaker-discriminative embeddings; (2) …
Transformers exhibit abrupt learning in matrix completion tasks.
problem Understanding abrupt learning in Transformers for matrix completion.
method Formulated matrix completion as MLM task, trained BERT model, analyzed model components.
result Sudden drop in loss despite no changes in training procedure or hyper-parameters.
DISTANA improves weather prediction by inferring hidden factors from temperature data.
problem Inferring hidden factors in spatiotemporal processes without supervision.
method Enhanced DISTANA architecture for spatiotemporal data, active tuning for latent state inference.
result DISTANA achieves more accurate predictions than other methods, inferring hidden factors from temperature data.
DOS improves language model generation by considering inter-token dependencies.
problem Lack of sequence-level information and inter-token dependencies in existing decoding strategies.
method Dependency-Oriented Sampler (DOS) that uses attention matrices to approximate inter-token dependencies.
result DOS consistently achieves superior performance on code generation and mathematical reasoning tasks.
Neural network based architectures used for sound recognition are usually adapted from other application domains, which may not harness sound related properties. The ConditionaL Neural Network (CLNN) is designed to consider the relational properties across frames in a temporal signal, and its extension the Masked Condi…
Recent variants improve knowledge distillation performance.
problem Improving the performance of knowledge distillation.
method Introducing additional components or changing the learning process.
result These variants have shown promising results.
Deep Neural Networks (DNNs) deliver state-of-the-art performance in many image recognition and understanding applications. However, despite their outstanding performance, these models are black-boxes and it is hard to understand how they make their decisions. Over the past few years, researchers have studied the proble…
Autoregressive models are among the best performing neural density estimators. We describe an approach for increasing the flexibility of an autoregressive model, based on modelling the random numbers that the model uses internally when generating data. By constructing a stack of autoregressive models, each modelling th…
PMI-Masking improves MLM pretraining by masking correlated spans efficiently.
problem Uniform token masking leads to inefficient and suboptimal performance in MLMs.
method PMI-Masking uses Pointwise Mutual Information to mask n-grams with high collocation.
result PMI-Masking reaches half the training time and improves performance.
A new method prunes neural network channels based on operation characteristics.
problem Compressing deep neural networks efficiently and maintaining accuracy.
method Differentiable masks for channel pruning considering BN and ReLU.
result Outstanding performance in accuracy with less resources compared to state-of-the-art methods.
A new method averages neural network parameters to rank features robustly.
problem Neural networks' sensitivity to random initialization affects feature ranking robustness.
method Parameter averaging of multiple shallow networks trained with different random seeds.
result The averaged model discovers ground-truth feature importance consistently.
The α-Alternator adapts to varying noise levels in sequences, improving robustness and performance.
problem Current models assume uniform noise levels, limiting performance on noisy temporal data.
method Introduces α-Alternator using Vendi Score to dynamically adjust noise sensitivity. result Outperforms Alternators and state-of-the-art models in trajectory prediction, imputation, and forecasting.
GIB improves neural network generalization by dynamically selecting task-relevant features across different sequential environments.
problem Poor generalization of deep neural networks to unseen environments.
method Proposes a gated information bottleneck (GIB) approach that dynamically drops spurious correlations and selects task-relevant features.
result GIB outperforms other IB approaches in adversarial robustness and OOD detection.
Our work connects parameter magnitudes and Hessian eigenspaces in deep neural nets.
problem Understanding the relationship between parameter magnitudes and Hessian curvature in deep learning models.
method Developed a matrix-free algorithm based on sketched SVDs to measure similarity between parameter masks and Hessian eigenspaces.
result Top Hessian eigenvectors tend to be concentrated around larger parameters, indicating a connection between parameter magnitudes and loss curvature.
Study examines effects of pruning techniques on deep learning models.
problem Understanding the impact of pruning methods on deep learning model structure and dynamics.
method Investigated differences in connectivity and learning dynamics of pruned models using various iterative pruning techniques.
result Emergence of structure in pruned models through magnitude-based unstructured pruning and weight rewinding.
ST-MTM models complex time series by decomposing and masking seasonal and trend components.
problem Forecasting complex time series with intricate temporal variations.
method Seasonal-Trend Decomposition with Masking and Contrastive Learning.
result ST-MTM achieves superior forecasting performance compared to existing methods.
Transformers' self-attention mechanism is mapped to a generalized Potts model.
problem Uncertainty in what type of data distribution self-attention can efficiently learn.
method Decouple word positions and embeddings, then show self-attention learns a generalized Potts model.
result Training self-attention is equivalent to solving the inverse Potts problem.
Expands MLM by masking token positions, improving performance and convergence.
problem Improving language model performance and convergence.
method Masking token positions along with [MASK] tokens, using a fully connected classifier stage.
result Shows .3% improvement and 50% faster convergence for BERT Base with position masking.
Neural networks with low-precision weights and activations offer compelling efficiency advantages over their full-precision equivalents. The two most frequently discussed benefits of quantization are reduced memory consumption, and a faster forward pass when implemented with efficient bitwise operations. We propose a t…
BetaExplainer improves GNN interpretability by masking unimportant edges.
problem Interpreting GNNs' predictions is difficult due to black-box behavior and lack of uncertainty quantification.
method BetaExplainer uses a sparsity-inducing prior to mask unimportant edges during training.
result BetaExplainer provides uncertainty in edge importance and improves predictive accuracy on challenging datasets.