VLMs Can Aggregate Scattered Training Patches

Whitepaper called VLMs Can Aggregate Scattered Training Patches...

FERRET: Private Deep Learning Faster and Better Than DPSGD

We revisit 1-bit gradient compression through the lens of mutual-information differential privacy MI-DP. Building on signSGD, we propose FERRET--Fast and Effective Restricted Release for Ethical Training--which transmits at most one sign bit per parameter group with Bernoulli masking. Theory: We...

Prediction Inconsistency Helps Achieve Generalizable Detection of Adversarial Examples

Adversarial detection protects models from adversarial attacks by refusing suspicious test samples. However, current detection methods often suffer from weak generalization: their effectiveness tends to degrade significantly when applied to adversarially trained models rather than naturally train...

Mind the Gap: a Practical Attack on GGUF Quantization

With the increasing size of frontier LLMs, post-training quantization has become the standard for memory-efficient deployment. Recent work has shown that basic rounding-based quantization schemes pose security risks, as they can be exploited to inject malicious behaviors into quantized models tha...

Robust Anti-Backdoor Instruction Tuning in LVLMs

Large visual language models LVLMs have demonstrated excellent instruction-following capabilities, yet remain vulnerable to stealthy backdoor attacks when finetuned using contaminated data. Existing backdoor defense techniques are usually developed for single-modal visual or language models under...

Which Factors Make Code LLMs More Vulnerable to Backdoor Attacks? A Systematic Study

Code LLMs are increasingly employed in software development. However, studies have shown that they are vulnerable to backdoor attacks: when a trigger a specific input pattern appears in the input, the backdoor will be activated and cause the model to generate malicious outputs. Researchers have...

Security Concerns for Large Language Models: a Survey

Large Language Models LLMs such as GPT-4 and its recent iterations, Google's Gemini, Anthropic's Claude 3 models, and xAI's Grok have caused a revolution in natural language processing, but their capabilities also introduce new security vulnerabilities. In this survey, we provide a comprehensive...

Towards Secure MLOps: Surveying Attacks, Mitigation Strategies, and Research Challenges

The rapid adoption of machine learning ML technologies has driven organizations across diverse sectors to seek efficient and reliable methods to accelerate model development-to-deployment. Machine Learning Operations MLOps has emerged as an integrative approach addressing these requirements by...

Hush! Protecting Secrets during Model Training: an Indistinguishability Approach

We consider the problem of secret protection, in which a business or organization wishes to train a model on their own data, while attempting to not leak secrets potentially contained in that data via the model. The standard method for training models to avoid memorization of secret information i...

Adversarial Threat Vectors and Risk Mitigation for Retrieval-Augmented Generation Systems

Retrieval-Augmented Generation RAG systems, which integrate Large Language Models LLMs with external knowledge sources, are vulnerable to a range of adversarial attack vectors. This paper examines the importance of RAG systems through recent industry adoption trends and identifies the prominent...

Adversarial Machine Learning for Robust Password Strength Estimation

Passwords remain one of the most common methods for securing sensitive data in the digital age. However, weak password choices continue to pose significant risks to data security and privacy. This study aims to solve the problem by focusing on developing robust password strength estimation models...

CVE-2025-5171

A vulnerability, which was classified as critical, has been found in llisoft MTA Maita Training System 4.5. This issue affects the function this.fileService.download of the file com\llisoft\controller\OpenController.java. The manipulation of the argument url leads to unrestricted upload. The atta...

CVE-2025-5171

A vulnerability, which was classified as critical, has been found in llisoft MTA Maita Training System 4.5. This issue affects the function this.fileService.download of the file com\llisoft\controller\OpenController.java. The manipulation of the argument url leads to unrestricted upload. The atta...

CVE-2025-5171

The CVE-2025-5171 entry concerns llisoft MTA Maita Training System 4.5. Affected: the file download path through this.fileService.download in com\llisoft\controller\OpenController.java. Root cause: argument url manipulation enables unrestricted upload. Impact: remote attack possible with high sev...

CVE-2025-5171 llisoft MTA Maita Training System OpenController.java this.fileService.download unrestricted upload

A vulnerability, which was classified as critical, has been found in llisoft MTA Maita Training System 4.5. This issue affects the function this.fileService.download of the file com\llisoft\controller\OpenController.java. The manipulation of the argument url leads to unrestricted upload. The atta...

CVE-2025-5170

A vulnerability classified as critical was found in llisoft MTA Maita Training System 4.5. This vulnerability affects the function AdminShitiListRequestVo of the file com\llisoft\controller\admin\shiti\AdminShitiController.java. The manipulation of the argument stTypeIds leads to sql injection. T...

CVE-2025-5170

The CVE-2025-5170 issue affects llisoft MTA Maita Training System version 4.5, specifically the AdminShitiListRequestVo function in com\llisoft\controller\admin\shiti\AdminShitiController.java. The vulnerability arises from improper handling of the stTypeIds argument, enabling SQL injection that ...

CVE-2025-5170 llisoft MTA Maita Training System AdminShitiController.java AdminShitiListRequestVo sql injection

A vulnerability classified as critical was found in llisoft MTA Maita Training System 4.5. This vulnerability affects the function AdminShitiListRequestVo of the file com\llisoft\controller\admin\shiti\AdminShitiController.java. The manipulation of the argument stTypeIds leads to sql injection. T...

Differential Privacy Analysis of Decentralized Gossip Averaging under Varying Threat Models

Fully decentralized training of machine learning models offers significant advantages in scalability, robustness, and fault tolerance. However, achieving differential privacy DP in such settings is challenging due to the absence of a central aggregator and varying trust assumptions among nodes. I...

llisoft MTA Maita Training System 代码问题漏洞

The llisoft MTA Maita Training System is a training system from China Dongke llisoft. A code issue exists in version 4.5 of the llisoft MTA Maita Training System, which is caused by a parameter url operation that results in unlimited uploads...