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.
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.
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.
Proposes a method to mask sensitive data while maintaining classifier usability and privacy.
problem Privacy-preserving data release with high usability.
method Data masking with privacy guarantees, ensuring classifier similarity.
result The proposed method achieves lower risk compared to input perturbation, especially with large sample sizes.
Detects adversarial examples using attribution methods.
problem Detecting adversarial examples in machine learning models.
method Integrated Gradient method for finding attributions and masking high attribution features to define causal neighborhoods.
result Adversarial inputs are not robust to masking high attribution features, while benign inputs are.
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.
Develops masks to explain neural network predictions.
problem Improving neural network interpretability for various applications.
method Creates explanation masks for pre-trained networks using a secondary network.
result Demonstrates the effectiveness of the method across different types of networks.
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).
VoiceFilter separates target speaker from multi-speaker signals.
problem Speech recognition in multi-speaker environments.
method Speaker recognition network and spectrogram masking network trained together.
result Significant reduction in speech recognition WER on multi-speaker signals.
Parallel decoding improves machine translation efficiency and accuracy.
problem Efficiently generating translations from left to right.
method Conditional masked language modeling for parallel decoding.
result Improves translation performance by over 4 BLEU points.
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.
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.
Paper proposes energy-efficient DNN training methods.
problem Energy-constrained deployment of deep neural networks.
method Weighted sparse projection and layer input masking integrated into DNN training.
result Framework provides higher accuracy with same or lower energy budgets.
New unsupervised neural network for beamforming masks trained on real noisy speech.
problem Training neural beamforming masks without labeled data.
method Unsupervised training using a likelihood criterion from spatial mixture model.
result Performance comparable to supervised and teacher-trained systems.
Modulating masks improve lifelong reinforcement learning.
problem Catastrophic forgetting and task interference in lifelong reinforcement learning.
method Adapted modulating masks for deep lifelong reinforcement learning (LRL) with PPO and IMPALA agents.
result Superior performance in both discrete and continuous RL tasks compared to LRL baselines.
AECF improves multimodal inference robustness and calibration.
problem Robustness and calibration issues in multimodal systems with missing inputs.
method Adaptive Entropy-Gated Contrastive Fusion (AECF) layer.
result Improves masked-input mAP by +18 pp at a 50% drop rate.
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.
ReDO extracts object masks from images without annotations.
problem Solving object segmentation with supervised learning requires large datasets and pixel-level annotations.
method ReDO uses an adversarial architecture where a generator redraws objects in images without annotations.
result ReDO successfully extracts high-quality object masks from various datasets.
Novel bio-inspired masking for robust speech emotion recognition.
problem Noise degradation in speech emotion recognition.
method Cochlear cepstrogram-based contrastive learning with temporal and frequency masking.
result Improved speech emotion recognition performance on K-EmoCon benchmark.
A new network predicts object/non-object masks to improve crowd counting.
problem Counting objects in images or videos with varying density.
method Proposes a network with a branch for object/non-object mask prediction, integrating it into the density map estimation.
result Demonstrates superior performance compared to baseline methods on five public datasets.
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.
Non-autoregressive transformer improves speech recognition speed and accuracy.
problem Reducing inference computation cost in speech recognition.
method Two non-autoregressive transformer structures (A-CMLM and A-FMLM) trained with masked tokens and iterative prediction during inference.
result Non-autoregressive transformer can match autoregressive transformer performance with 7x speedup.
Mask-reconstruction pretraining helps in downstream tasks by capturing more semantic features.
problem How mask-reconstruction pretraining helps in downstream tasks and why it surpasses supervised learning.
method Theoretical analysis and experimental validation of mask-reconstruction pretraining (MRP) on auto-encoders.
result MRP provably captures more semantic features than supervised learning, leading to better performance in downstream tasks.
Machine learning predicts mask mandates reduce COVID-19 deaths.
problem Effectiveness of mask mandates on reducing COVID-19 deaths.
method Machine learning classification algorithms applied to survey data.
result Mask mandates correlated with reduced death ratios in most counties.
CNNs overinterpret inputs, leading to high accuracy without meaningful features.
problem High accuracy in image classifiers can mask subtle model failures.
method Batched Gradient SIS method for discovering sufficient input subsets.
result Overinterpretation allows models to make confident predictions with masked input features.
RGI improves robustness of GAN-inversion for image restoration and anomaly detection.
problem Robustness of GAN-inversion to unknown gross corruptions.
method Proposes RGI and R-RGI methods with provable robustness guarantees.
result Restored images and corrupted region masks converge to ground truth under mild assumptions.
The paper analyzes CFG in masked diffusion models and proposes a new method to improve sample quality.
problem Improving sample quality in conditional generation with CFG in masked diffusion models.
method Analyzing CFG's effect in a low-dimensional masked diffusion model, revealing early-stage guidance harms quality and late-stage improves it. Proposed a novel method to smooth transitions between distributions.
result The proposed method improves sample quality via a simple one-line code change.
A new deep neural network improves mammogram image processing.
problem Improving mammogram image processing accuracy and efficiency.
method A novel deep neural network architecture with dual-path connections.
result Achieves best mammography segmentation and classification results.
There has been a lot of recent interest in designing neural network models to estimate a distribution from a set of examples. We introduce a simple modification for autoencoder neural networks that yields powerful generative models. Our method masks the autoencoder's parameters to respect autoregressive constraints: ea…
PSI models and infers feature attributions efficiently and accurately.
problem Modeling and inferring feature attributions in flexible predictive models.
method Probabilistic Shapley inference (PSI) framework using latent random variables and a masking-based neural network architecture.
result PSI learns feature attribution distributions centered at Shapley values, revealing meaningful uncertainty.
Reconstruction-based learning produces uninformative features for perception tasks.
problem Misalignment between reconstruction-based learning and perception tasks.
method Investigated the impact of input space reconstruction on feature learning for perception tasks.
result Reconstruction-based learning allocates model capacity to a subspace with uninformative features for perception tasks.
A reinforcement learning method predicts medical outcomes from masked feature vectors.
problem Predicting medical outcomes from limited feature data.
method Reinforcement learning with iterative feature selection and a guesser network.
result The method outperforms baselines and is highly interpretable.
This paper explores four different visualization techniques for long short-term memory (LSTM) networks applied to continuous-valued time series. On the datasets analysed, we find that the best visualization technique is to learn an input deletion mask that optimally reduces the true class score. With a specific focus o…
NeuroMask provides interpretable explanations for deep neural networks.
problem Understanding how deep neural networks make decisions.
method Applies a mask to reveal or hide parts of an image, tuning mask values to preserve classification results and produce interpretable explanations.
result NeuroMask successfully localizes the most relevant parts of an image to a deep neural network's decision.
New input gradient regularization improves adversarial robustness efficiently.
problem Improving adversarial robustness in machine learning models.
method Derive robustness bounds, implement scaleable input gradient regularization, avoid double backpropagation.
result Input gradient regularization is competitive with adversarial training and avoids gradient obfuscation.
SpecAugment improves speech recognition with simple feature augmentation.
problem Improving automatic speech recognition accuracy.
method Applying warping, frequency channel masking, and time step masking to feature inputs of neural networks.
result Achieved state-of-the-art performance on LibriSpeech and Switchboard tasks.
A framework to explain decoder-only sequence classification models using intermediate predictions.
problem Explaining predictions of decoder-only sequence classification models.
method Progressive Inference framework with Single Pass-Progressive Inference and Multi Pass-Progressive Inference methods.
result Significantly better attributions compared to prior work on text classification tasks.
This work improves explanation quality for time series predictions by learning perturbations.
problem Explaining predictions on multivariate time series data with time dependencies.
method Learning both masks and associated perturbations to explain predictions.
result Learning perturbations significantly improves explanation quality on time series data.
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.
In this work we develop a fast saliency detection method that can be applied to any differentiable image classifier. We train a masking model to manipulate the scores of the classifier by masking salient parts of the input image. Our model generalises well to unseen images and requires a single forward pass to perform …
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.
The paper offers generalization bounds for Transformers that ignore sequence length.
problem Developing generalization bounds for Transformers that are independent of sequence length.
method Covering number approach to upper bound Rademacher complexity of bounded linear transformations.
result Theoretical bounds for Transformer generalization are independent of sequence length.
LcGAN generates synthetic CT images for hemorrhagic lesion segmentation.
problem Scarce training data for hemorrhagic lesion segmentation.
method Lesion conditional Generative Adversarial Network (LcGAN) for synthetic image generation.
result Segmentation improved by 12.8% with synthetic data augmentation.
Paper uses GMM and MAF for probabilistic classification, outperforming simpler models.
problem Classifying data with complex distributions.
method Density estimation using Gaussian Mixture Model and Masked Autoregressive Flow.
result Proposed classifiers outperform simpler models like linear discriminant analysis.
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.
AaSP improves audio self-supervised learning by addressing aliasing issues.
problem Alias issues in audio spectrogram transformers.
method AaSP combines aliasing-aware patch representation, teacher-student masked modeling, cross-attention predictor, and contrastive regularization.
result AaSP learns more stable representations that integrate high-frequency cues.
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.
Proposes QNN to protect input privacy in neural networks.
problem Protecting input privacy in neural networks.
method Quaternion-valued neural network (QNN) to hide input information.
result QNN effectively protects input privacy without significant accuracy loss.