LAUSR

Membership Inference Attacks With Token-Level Deduplication on Korean Language Models

Sign Up to like & get
recommendations!
2 Published in 2023 at "IEEE Access"

DOI: 10.1109/access.2023.3239668

Abstract: The confidentiality threat against training data has become a significant security problem in neural language models. Recent studies have shown that memorized training data can be extracted by injecting well-chosen prompts into generative language models.… read more here.

Keywords: token level; language; attack; language models ... See more keywords

Label-Only Membership Inference Attacks and Defenses in Semantic Segmentation Models

Sign Up to like & get
recommendations!
2 Published in 2023 at "IEEE Transactions on Dependable and Secure Computing"

DOI: 10.1109/tdsc.2022.3154029

Abstract: Recent research has discovered that deep learning models are vulnerable to membership inference attacks, which can reveal whether a sample is in the training dataset of the victim model or not. Most membership inference attacks… read more here.

Keywords: membership inference; inference attacks; segmentation models;

Practical Membership Inference Attack Against Collaborative Inference in Industrial IoT

Sign Up to like & get
recommendations!
1 Published in 2022 at "IEEE Transactions on Industrial Informatics"

DOI: 10.1109/tii.2020.3046648

Abstract: The effectiveness of state-of-the-art deep learning (DL) models has empowered the development of industrial Internet of things (IIoT). Recently, considering resource-constrained and privacy-required IIoT devices, collaborative inference has been proposed, which splits DL models and… read more here.

Keywords: inference attack; membership inference; inference; collaborative inference ... See more keywords

LTU Attacker for Membership Inference

Sign Up to like & get
recommendations!
1 Published in 2022 at "Algorithms"

DOI: 10.3390/a15070254

Abstract: We address the problem of defending predictive models, such as machine learning classifiers (Defender models), against membership inference attacks, in both the black-box and white-box setting, when the trainer and the trained model are publicly… read more here.

Keywords: ltu attacker; attacker; privacy; attack ... See more keywords

Enabling Trade-offs in Privacy and Utility in Genomic Data Beacons and Summary Statistics

Sign Up to like & get
recommendations!
2 Published in 2023 at "Genome research"

DOI: 10.48550/arxiv.2302.01763

Abstract: The collection and sharing of genomic data are becoming increasingly commonplace in research, clinical, and direct-to-consumer settings. The computational protocols typically adopted to protect individual privacy include sharing summary statistics, such as allele frequencies, or… read more here.

Keywords: privacy utility; genomic data; membership inference; summary statistics ... See more keywords