Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,932 papers · 148 categories

Trend · papers per month

4108211,2311,641 · Jun 202019922001200920172026
48 results for Deep Leakage from Gradient

Deep Leakage from Gradients can reveal private training data from shared gradients.

problem The safety of gradients in multi-node machine learning systems is questionable.
method Empirical validation of Deep Leakage from Gradient attack on computer vision and natural language processing tasks.
result The attack is more effective than previous methods, achieving pixel-wise accuracy for images and token-wise matching for texts.

Improved DLG extracts accurate labels from gradients, overcoming DLG's convergence issues.

problem Private training data leakage from shared gradients in distributed learning systems.
method Proposes iDLG, a simple approach to synthesize accurate labels from gradients.
result iDLG reliably extracts ground-truth labels from gradients, unlike DLG.

This paper analyzes how training data can be leaked from gradients in neural networks and proposes a metric for measuring model security.

problem Training data leakage from gradients in neural networks for image classification.
method Formulated the problem as an optimisation problem for each layer, involving weights, gradients, and constraints from preceding layers.
result Attributed training data leakage to the architecture of the deep network and proposed a metric for measuring model security.

This paper evaluates and compares gradient leakage attacks in federated learning.

problem Gradient leakage attacks compromise client privacy in federated learning.
method Formal and experimental analysis of gradient leakage attacks, evaluation of attack effectiveness and cost.
result Gradient leakage attacks can reconstruct private local training data from shared parameter updates.

Reduces data leakage in distributed deep learning models.

problem Prevents reconstruction of sensitive raw data patterns during client communications.
method Reduces distance correlation between raw data and learned representations.
result Resilient to reconstruction attacks while maintaining model accuracy.

Optimal defenses protect FL models from gradient reconstruction attacks.

problem Gradient reconstruction attacks compromise FL models by recovering original data from shared gradients.
method Derive a theoretical lower bound of reconstruction error, customize noise and pruning defenses, and achieve optimal trade-off between leakage and utility.
result Our methods outperform Gradient Noise and Pruning in protecting training data and maintaining model utility.

We survey distributed deep learning models for training or inference without accessing raw data from clients. These methods aim to protect confidential patterns in data while still allowing servers to train models. The distributed deep learning methods of federated learning, split learning and large batch stochastic gr…

2018-12-08abs ↗pdf ↗

HCBM improves deep learning explainability by non-linear concept aggregation.

problem Lack of explainable and accurate predictions in deep learning for high-stake decisions.
method Introduce Hoeffding Concept Bottleneck Models (HCBM) using Hoeffding functional decomposition of gradient-boosted trees for non-linear and sparse concept aggregation.
result HCBM outperforms standard linear CBM and is robust to interconcept leakage.

Study on privacy leakage in noisy gradient descent algorithms.

problem Information leakage of iterative randomized learning algorithms about training data.
method Analyzes the dynamics of Rényi differential privacy loss in noisy gradient descent algorithms.
result Privacy loss converges exponentially fast for smooth and strongly convex loss functions.

New model quantifies how much machine learning models can reveal about individual data usage.

problem Measuring and reducing the leakage of membership information from machine learning models.
method Using information theory, conditional mutual information leakage, and Kullback-Leibler divergence to quantify and bound the leakage.
result The amount of membership information leakage is reduced by adding Gaussian (ε,δ)(ε,δ)-differentially-private additive noises.

Underwater gas reservoirs are used in many situations. In particular, Carbon Capture and Storage (CCS) facilities that are currently being developed intend to store greenhouse gases inside geological formations in the deep sea. In these formations, however, the gas might percolate, leaking back to the water and eventua…

2019-04-11abs ↗pdf ↗

This paper bounds min-entropy leakage for Blowfish privacy using graph symmetries.

problem Bounding min-entropy leakage for Blowfish privacy mechanisms.
method Organizing analysis over symmetrical partitions corresponding to orbits of graph automorphism groups.
result Demonstrates a construction meeting the bound with asymptotic equality, showing tightness.

A deep learning framework learns wavelet packet transforms for efficient feature extraction.

problem Efficiently extracting meaningful time-frequency features from high-frequency signals.
method Learnable wavelet packet transforms using deep learning.
result Improved spectral leakage and enhanced anomaly detection performance.

This study examines how model architecture affects deep learning model privacy.

problem Privacy concerns in deep learning models due to potential leakage of sensitive information.
method Investigation of CNNs and Transformers, focusing on activation layers, stem layers, LN layers, and attention modules.
result Transformers generally exhibit higher vulnerability to privacy attacks than CNNs.

bioLeak addresses data leakage in biomedical machine learning studies.

problem Data leakage causes optimistic bias in machine learning models for biomedical studies.
method bioLeak provides leakage-aware resampling workflows and model audits in R.
result The package supports various machine learning tasks and can detect leakage mechanisms.

The paper analyzes privacy leakage in federated learning using linear algebra and optimization theory.

problem Privacy leakage in federated learning despite its promise for data privacy.
method Theoretical analysis from linear algebra and optimization theory perspectives.
result Derives sufficient conditions to prevent data reconstruction attacks and establishes an upper bound on privacy leakage.

Unified analysis of privacy leakage in correlated data considering prior knowledge.

problem Understanding the impact of prior knowledge and data correlation on privacy leakage.
method Proposed prior differential privacy (PDP) and analyzed using WHG and multivariate Gaussian models.
result Derived closed-form expression for continuous data and chain rule for discrete data.

Benchmark detects decision-time leakage in financial backtests.

problem Detecting decision-time leakage in financial machine-learning backtests.
method Toggles one evaluation convention at a time around a clean t+1t{+}1-open reference, holding other factors fixed.
result Inflation is highly selective, affecting specific features and execution methods.

Split learning preserves privacy in 1D CNN models for detecting heart abnormalities.

problem Privacy leakage in 1D CNN models under split learning.
method Implemented and validated an 1D CNN model under split learning, applied privacy leakage mitigation techniques.
result Split learning alone is insufficient to maintain raw data privacy in 1D CNN models.

Paper tackles treatment leakage in text-based causal inference, proposing methods to mitigate bias.

problem Treatment leakage in text-as-confounder applications introduces bias in causal estimates.
method Formal definitions, four text distillation methods (passage removal, classification, salient feature removal, nullspace projection).
result Moderate distillation optimally balances bias reduction against confounder retention.

Unified framework analyzes privacy risks from gradients in distributed learning.

problem Analyzing inference privacy risks from gradients in machine learning.
method Unified game-based framework for various attacks, including attribute, property, distributional, and user disclosures.
result Demonstrates inefficacy of data aggregation for privacy against inference attacks.

DPNR preserves privacy of text representations using differential privacy.

problem Privacy leakage in deep learning text representations.
method DPNR uses Differential Privacy to provide formal privacy guarantees and dropout masking for enhanced privacy.
result DPNR reduces privacy leakage without significantly sacrificing main task performance.

Improved privacy analysis for stochastic gradient descent.

problem Analyzing privacy leakage in noisy stochastic gradient descent.
method Modeling Rényi divergence dynamics with Langevin diffusions, proving exponential privacy loss convergence for smooth and strongly convex objectives.
result Privacy loss converges exponentially fast for smooth and strongly convex objectives under constant step size.

PPGAN adds noise to GAN gradients to protect private data.

problem Protecting private data in GANs when training on sensitive datasets.
method Integrates differential privacy into GANs by adding noise to gradients and using Moments Accountant strategy.
result Demonstrates generation of high-quality synthetic data while maintaining privacy.

We consider visual domains in which a class label specifies the content of an image, and class-irrelevant properties that differentiate instances constitute the style. We present a domain-independent method that permits the open-ended recombination of style of one image with the content of another. Open ended simply me…

2018-09-28abs ↗pdf ↗

Paper explores GAN generalization via privacy protection, proving bounds on generalization gap and reducing leakage.

problem Understanding and bounding the generalization gap of GANs.
method Theoretical proof using differential privacy, reinterpreting Bayesian GAN, membership attacks.
result Proven bounds on generalization gap and reduced information leakage.

Regularization may not fully protect against membership inference attacks.

problem Mitigating membership inference attacks in deep learning models.
method Evaluation of 8 regularization mechanisms on three image classification tasks.
result Certain regularization mechanisms can inadvertently help membership inference attacks.

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.