Latest Articles for Cybersecurity

A new framework enhances language model security against harmful data manipulation.

A look at new methods to combat data poisoning attacks in machine learning.

Instruction tuning raises concerns over security vulnerabilities in large language models.

A new model enhances the detection of fake audio clips through stereo analysis.

A new method improves privacy in decision trees without sacrificing accuracy.

Exploring the risks of large language models in cyber threats.

Assessing AI models is vital to prevent harmful outcomes in technology development.

LLMDet identifies the source of text, enhancing trust in information.

A new method improves realism in adversarial samples while maintaining effectiveness.

Radio fingerprinting enhances satellite security against spoofing and replay attacks.

Exploring methods to detect fraud in DeFi using advanced data analysis techniques.

This study examines attacks on water distribution networks and proposes methods for better leak detection.

A method to enhance text classification models against adversarial changes.

A new model enhances detection of cyber threats in cloud environments.

A new method enhances anonymity in ring signatures for better security.

Examining the role of FPGA techniques in enhancing hardware security.

A new method predicts network robustness efficiently using deep learning techniques.

A new model enhances vulnerability detection accuracy using deep learning techniques.

A look at new encoding methods for better data privacy.

MeTTa aims to unify programming for humans and AGIs.

Assessing privacy risks in deep learning against various attacks.

A new protocol to protect user privacy in 5G network slicing.

PPA-AFL improves data privacy in collaborative machine learning across devices.

A new method for detecting unusual patterns in data using machine learning.

Incremental learning enhances machine learning but faces serious threats from backdoor attacks.

This study explores how network operators adapted during the Russia-Ukraine conflict.

A deep dive into face morphing attacks and detection methods.

UMD offers a new way to identify complex backdoor attacks effectively.

A study on the effectiveness of neighborhood attacks in revealing training data.

A new framework enhances the detection of cyber attack paths.

T800 offers an efficient security solution for vulnerable IoT devices.

Explore strategies to ensure safe routing in vulnerable networks.

A new method enhances detection of ransomware using multiple tests for better accuracy.

A look at the risks and vulnerabilities of fingerprint security on smartphones.

A new method leveraging graphs to identify adversarial attacks on neural networks.

This research explores how adversarial attacks affect Bayesian inference models.

Exploring clean-label attacks and defenses in NLP machine learning models.

Examining how subtle changes impact image quality in compression systems.

This review looks into the resilience of unsupervised learning to adversarial attacks.

This article examines attribute inference attacks on GNNs and their impact on privacy.