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.
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.
The paper investigates the effects of invalid action masking in policy gradient algorithms.
problem Invalid actions in policy gradient algorithms can lead to suboptimal performance.
method The paper provides theoretical justification and empirical demonstrations of the importance of invalid action masking.
result Invalid action masking is crucial as the number of invalid actions increases.
A new framework for controllable generation of discrete masked models.
problem Efficient controllable generation of discrete data models.
method Plug-and-play framework based on importance sampling.
result Demonstrates versatility across multiple domains, including protein design.
Sparse random networks reduce communication in federated learning.
problem Large communication cost in federated learning.
method Freeze random weights, train stochastic binary mask to sparsify.
result Improves accuracy, reduces communication, speeds convergence.
CACTI improves tabular data imputation by leveraging missingness patterns and contextual information.
problem Tabular data imputation with improved accuracy and robustness.
method Masked autoencoding approach with median truncated copy masking and contextual information.
result Average R2 gain of 7.8% over the next best method across various datasets and missingness conditions. 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.
A new method for graph-structured data improves transformer performance by incorporating topology.
problem Improving transformer performance on graph-structured data.
method Parameterizing topological masks as a learnable function of a weighted adjacency matrix, approximated with graph random features.
result Efficient masking algorithms provide strong performance gains for tasks on image and point cloud data.
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.
Study improves healthcare time series imputation by considering structured missingness.
problem Structured missingness in clinical data impacts time series imputation models.
method Analysis of different masking strategies on imputation methods using PhysioNet Challenge 2012 dataset.
result Masking choices significantly affect imputation accuracy and clinical prediction.
Proposes a proportional masking strategy for better tabular data imputation.
problem Heterogeneity of tabular data disrupts uniform random masking in MAEs.
method Computes missingness statistics, generates proportional masks, uses MLP token mixing.
result Proportional masking preserves missingness distribution, improves imputation performance.
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…
This paper connects masked pre-training to Bayesian model selection.
problem Understanding the success of masked pre-training and its generalization.
method The paper shows masked pre-training corresponds to maximizing the marginal likelihood.
result Masked pre-training with a suitable scoring function maximizes the marginal likelihood.
A deep generative model improves imputation of MNAR data by treating missing and complete data equally.
problem Missing Not At Random (MNAR) data in analysis.
method A generative model-specific joint probability decomposition method (conjunction model) and a deep generative imputation model (GNR).
result GNR surpasses state-of-the-art MNAR baselines with significant margins in RMSE and better mask reconstruction.
New analysis reveals masked self-supervised learning's effectiveness in extracting data structure.
problem Analyzing masked self-supervised learning in high-dimensional data.
method Developed precise high-dimensional analysis of masked modeling objectives.
result Identified phase transitions and structured regimes for masked self-supervised learning.
Square Attack efficiently attacks deep models with random updates.
problem Efficiently attacking deep learning models without gradient information.
method Randomized search of localized square updates.
result Significantly more query-efficient and higher success rate compared to state-of-the-art methods.
We introduce backdrop, a flexible and simple-to-implement method, intuitively described as dropout acting only along the backpropagation pipeline. Backdrop is implemented via one or more masking layers which are inserted at specific points along the network. Each backdrop masking layer acts as the identity in the forwa…
Strong bolicity helps prove Baum-Connes conjecture for certain hyperbolic groups.
problem Proving the Baum-Connes conjecture for relatively hyperbolic groups.
method Constructing a strongly bolic metric and using masks for random coset representatives.
result Deduced the Baum-Connes conjecture for groups satisfying (RD) and certain parabolics.
New taxonomy divides defense methods for neural networks.
problem Improving adversarial robustness of neural networks.
method Reframing existing defense categories into two new categories.
result There is no universal trade-off between robustness and accuracy.
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.
TRM improves long-horizon LLM RL by masking divergent sequences.
problem Long-horizon reinforcement learning with LLMs suffers from off-policy mismatch and approximation errors.
method Derives and applies trust region bounds to control divergence, proposing Trust Region Masking.
result First non-vacuous monotonic improvement guarantees for long-horizon LLM-RL.
TRM improves long-horizon reinforcement learning for LLMs by masking divergent sequences.
problem Long-horizon reinforcement learning for LLMs suffers from off-policy mismatch and approximation errors.
method Derives and applies trust region bounds to control divergence, proposing Trust Region Masking.
result First non-vacuous monotonic improvement guarantees for long-horizon LLM-RL.
MCSAE improves speaker embedding by focusing on both high- and low-level features.
problem Reduced effect of low-level features in speaker embedding encoding.
method Masked cross self-attentive encoding using ResNet with multi-layer aggregation and random masking regularization.
result Improved speaker embedding with equal error rate of 2.63% and minimum detection cost function of 0.1453.
Method generates anatomically-controllable medical images with segmentation guidance.
problem Challenging to enforce anatomical constraints in generated medical images.
method Segmentation-guided diffusion models with random mask ablation training.
result New state-of-the-art in faithfulness to input anatomical masks.
MDMs train to decode tokens in a random order, which affects performance; we show they can be optimized for a favorable order.
problem Performance of MDMs is affected by the random order in which tokens are decoded.
method We show that MDMs can be optimized for a favorable decoding order by equipping their continuous-time variational objective with multivariate noise schedules.
result MDMs can be decomposed into a weighted auto-regressive losses over orders, making them auto-regressive models with learnable orders.
In this work we revisit gradient regularization for adversarial robustness with some new ingredients. First, we derive new per-image theoretical robustness bounds based on local gradient information. These bounds strongly motivate input gradient regularization. Second, we implement a scaleable version of input gradient…
A new approach to protein language models combines latent space prediction with masked language modeling.
problem Improving protein language models by predicting amino acid identities at masked positions.
method A variant of masked language modeling that predicts latent targets only at masked positions, retaining the MLM cross-entropy.
result The new approach outperforms pure masked language modeling on 11 out of 16 downstream tasks.
LightSecAgg reduces secure aggregation complexity in FL.
problem Complexity in secure aggregation protocols for FL systems.
method One-shot aggregate-mask reconstruction via mask encoding/decoding.
result Significantly reduces overhead for resiliency against dropped users.
New criteria improve imputation model selection using MOO.
problem Selecting the best imputation model using prediction accuracy metrics.
method Introduced three modified MOO criteria based on rank transformation, energy distance, and likelihood principle.
result Demonstrated how MOO is related to missing-at-random assumption and derived statistical and computational learning theories.
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.
Develops a multilevel Monte Carlo framework with dropout for efficient uncertainty quantification.
problem Efficiently quantify uncertainty in complex models using dropout.
method Integrates multilevel Monte Carlo with Monte Carlo dropout, creating coupled estimators to reduce variance.
result Demonstrates significant variance reduction and efficiency gains over single-level Monte Carlo dropout.
SupSup model learns thousands of tasks without forgetting, using randomly initialized subnetworks.
problem Sequentially learning many tasks without forgetting.
method Randomly initialized base network with task-specific subnetworks (supermasks).
result Gradient-based optimization can identify the correct subnetwork for new tasks.
This paper explains CART random forests using stochastic control theory.
problem Understanding the inner workings of CART random forests.
method Developed a stochastic-control perspective on CART random forests, interpreting feature subsampling as a random feasible action set and the split rule as a policy.
result Established that the CART policy is locally stabilizing but globally suboptimal for the forest objective.
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).
FMix improves model performance by distorting learned functions without distorting data distribution.
problem Improving model robustness and generalization performance in mixed sample data augmentation.
method Proposes FMix, an MSDA using random binary masks generated from Fourier space.
result FMix outperforms MixUp and CutMix, achieving state-of-the-art results on CIFAR-10.
Improves generative models by optimizing rewards and sample editing.
problem Efficiently generating high-reward samples with structural constraints.
method Introduces MDM-VGB, a discrete diffusion sampler that augments unmasking generation with reward-guided remasking.
result MDM-VGB achieves quadratic complexity and robustness to noise, outperforming heuristics like best-of-N. Masked LARk prevents cross-site tracking while training models.
problem Cross-site tracking of user data through third-party cookies.
method Secure multi-party compute (MPC) protocol with masking.
result Prevents cross-site tracking and maintains model training flexibility.
Secure federated learning framework resists adversarial users.
problem Resilience against adversarial (Byzantine) users in federated learning.
method Integrated stochastic quantization, verifiable outlier detection, and secure model aggregation.
result First single-server Byzantine-resilient secure aggregation framework (BREA) for secure federated learning.
The kernel embedding algorithm is an important component for adapting kernel methods to large datasets. Since the algorithm consumes a major computation cost in the testing phase, we propose a novel teacher-learner framework of learning computation-efficient kernel embeddings from specific data. In the framework, the h…
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…
Attribution methods have been developed to explain the decision of a machine learning model on a given input. We use the Integrated Gradient method for finding attributions to define the causal neighborhood of an input by incrementally masking high attribution features. We study the robustness of machine learning model…
A key problem in research on adversarial examples is that vulnerability to adversarial examples is usually measured by running attack algorithms. Because the attack algorithms are not optimal, the attack algorithms are prone to overestimating the size of perturbation needed to fool the target model. In other words, the…
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.
Quantized neural networks are vulnerable to adversarial attacks.
problem Adversarial robustness of quantized neural networks.
method Investigated adversarial robustness of quantized neural networks under different threat models.
result Quantization does not offer robust protection and results in gradient masking.
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.
SMART training improves mask-predict translations.
problem Closing the performance gap between semi-autoregressive and autoregressive models.
method SMART training method for conditional masked language models.
result SMART-trained models produce higher-quality translations.
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.
PLS-SVD struggles with missing data in multimodal datasets, showing a phase transition in performance.
problem Missing data in PLS-SVD for multimodal datasets.
method Replica-symmetric analysis of spiked rectangular random matrices with missing entries.
result PLS-SVD performance transitions from uninformative to informative singular vectors at a critical signal-to-noise threshold.