DP-SGD
PulseAugur coverage of DP-SGD — every cluster mentioning DP-SGD across labs, papers, and developer communities, ranked by signal.
9 day(s) with sentiment data
-
New DP learning framework uses hypernetwork to reduce noise impact
Researchers have developed a novel framework for differentially private (DP) learning that bypasses iterative parameter-space optimization. Instead of using privatized gradients, the method employs a hypernetwork traine…
-
New auditors improve f-Differential Privacy assessment without fixed sample size
Researchers have developed new auditors to empirically assess the Differential Privacy (DP) of algorithms, focusing on the expressive $f$-DP concept. These auditors can detect privacy violations across the full privacy …
-
New papers explore differential privacy in Gaussian Processes and ML reporting
Two recent arXiv papers explore differential privacy in machine learning, focusing on Gaussian processes and reporting mechanisms. The first paper details how the intrinsic randomness of Gaussian Process posterior sampl…
-
New TRAP benchmark reveals AI agents leak sensitive data, proposes isolation solution · 3 sources tracked
Researchers have introduced TRAP, a new benchmark designed to evaluate AI agents' ability to complete tasks while resisting privacy extraction. The benchmark assesses the trade-off between task accuracy and data leakage…
-
Federated autoencoder enhances ECG anomaly detection with privacy on edge devices
Researchers have developed a privacy-preserving federated autoencoder system for detecting anomalies in electrocardiogram (ECG) data on edge devices. The system combines federated learning with differential privacy and …
-
New research papers explore robust privacy and differential privacy in ML
Two new research papers explore advanced privacy techniques for machine learning models. The first paper introduces "Robust Privacy" (RP), a method that leverages certified robustness to protect sensitive attributes dur…
-
EEG Foundation Models Leak Data Despite Standard Audits
Researchers have developed a new auditing framework for EEG foundation models that goes beyond single-endpoint evaluations. This framework jointly audits multiple endpoints, revealing that models cleared by individual t…
-
New DP-SGD method updates fewer coordinates for efficiency
Researchers have developed a new method called TP-TopK DP-SGD to improve the efficiency of differentially private stochastic gradient descent. This technique aims to reduce the computational overhead by updating fewer c…
-
New PRISM method enhances differential privacy for LoRA models
Researchers have introduced PRISM, a novel method for applying differential privacy to Low-Rank Adaptation (LoRA) in machine learning models. Traditional methods struggle because LoRA's low-rank factorization is not uni…
-
New Research Links Privacy and Generalization in DP-SGD
A new research paper titled "From Privacy to Generalization: Linear Max-Information Bounds for DP-SGD" has been published on arXiv. The paper addresses the challenge of understanding the link between generalization and …
-
New framework offers optimal guarantees for auditing RDP machine learning
Researchers have developed a new auditing framework for machine learning algorithms that claim Rényi differential privacy (RDP). This framework uses the Donsker-Varadhan (DV) estimator to directly measure Rényi divergen…
-
New theory bounds KAN training, reveals privacy-utility gap
Researchers have established new theoretical bounds for training Kolmogorov-Arnold Networks (KANs), a structured alternative to standard MLPs. The work analyzes KANs trained with mini-batch stochastic gradient descent (…
-
New DP-LAC method enhances private federated LLM fine-tuning
Researchers have developed DP-LAC, a new method for differentially private federated fine-tuning of language models. This technique improves upon existing adaptive clipping methods by estimating an initial clipping thre…
-
New DP-SGD subsampling methods offer improved privacy-utility trade-offs
Two new research papers explore optimized subsampling techniques for Differentially Private Stochastic Gradient Descent (DP-SGD). The first paper, focusing on random shuffling, provides tight upper and lower bounds with…
-
Researchers reveal supply-chain attacks can steal secrets from local LLM fine-tuning
Researchers have developed a novel method to steal sensitive information from locally fine-tuned large language models by exploiting vulnerabilities in their supply chain code. This technique moves beyond passive weight…
-
Google Research unveils private synthetic photo album generation
Google Research has developed a novel method for generating differentially private synthetic photo albums. This approach utilizes an intermediate text representation and a hierarchical generation process to ensure data …