Fragments to Facts: Partial-Information Fragment Inference from LLMs

Large language models LLMs can leak sensitive training data through memorization and membership inference attacks. Prior work has primarily focused on strong adversarial assumptions, including attacker access to entire samples or long, ordered prefixes, leaving open the question of how vulnerable...

MorphMark: Flexible Adaptive Watermarking for Large Language Models

Watermarking by altering token sampling probabilities based on red-green list is a promising method for tracing the origin of text generated by large language models LLMs. However, existing watermark methods often struggle with a fundamental dilemma: improving watermark effectiveness the...

R1dacted: Investigating Local Censorship in DeepSeek'S R1 Language Model

DeepSeek recently released R1, a high-performing large language model LLM optimized for reasoning tasks. Despite its efficient training pipeline, R1 achieves competitive performance, even surpassing leading reasoning models like OpenAI's o1 on several benchmarks. However, emerging reports suggest...

Benchmarking LLMs in an Embodied Environment for Blue Team Threat Hunting

As cyber threats continue to grow in scale and sophistication, blue team defenders increasingly require advanced tools to proactively detect and mitigate risks. Large Language Models LLMs offer promising capabilities for enhancing threat analysis. However, their effectiveness in real-world blue...

FABLE: a Localized, Targeted Adversarial Attack on Weather Forecasting Models

Deep learning-based weather forecasting models have recently demonstrated significant performance improvements over gold-standard physics-based simulation tools. However, these models are vulnerable to adversarial attacks, which raises concerns about their trustworthiness. In this paper, we first...

TechniqueRAG: Retrieval Augmented Generation for Adversarial Technique Annotation in Cyber Threat Intelligence Text

Accurately identifying adversarial techniques in security texts is critical for effective cyber defense. However, existing methods face a fundamental trade-off: they either rely on generic models with limited domain precision or require resource-intensive pipelines that depend on large labeled...

On Membership Inference Attacks in Knowledge Distillation

Nowadays, Large Language Models LLMs are trained on huge datasets, some including sensitive information. This poses a serious privacy concern because privacy attacks such as Membership Inference Attacks MIAs may detect this sensitive information. While knowledge distillation compresses LLMs into...

CVE-2024-56427

An issue was discovered in Samsung Mobile Processor and Wearable Processor Exynos 980, 990, 850, 1080, 2100, 1280, 2200, 1330, 1380, 1480, 2400, 9110, W920, W930, W1000, Modem 5123, Modem 5300, Modem 5400. The lack of a length check leads to out-of-bounds access via malformed RRC packets to the...

ProxyPrompt: Securing System Prompts against Prompt Extraction Attacks

The integration of large language models LLMs into a wide range of applications has highlighted the critical role of well-crafted system prompts, which require extensive testing and domain expertise. These prompts enhance task performance but may also encode sensitive information and filtering...

GenoArmory: a Unified Evaluation Framework for Adversarial Attacks on Genomic Foundation Models

We propose the first unified adversarial attack benchmark for Genomic Foundation Models GFMs, named GenoArmory. Unlike existing GFM benchmarks, GenoArmory offers the first comprehensive evaluation framework to systematically assess the vulnerability of GFMs to adversarial attacks. Methodologicall...

AutoRAN: Weak-To-Strong Jailbreaking of Large Reasoning Models

This paper presents AutoRAN, the first automated, weak-to-strong jailbreak attack framework targeting large reasoning models LRMs. At its core, AutoRAN leverages a weak, less-aligned reasoning model to simulate the target model's high-level reasoning structures, generates narrative prompts, and...

Diverging Towards Hallucination: Detection of Failures in Vision-Language Models Via Multi-Token Aggregation

Vision-language models VLMs now rival human performance on many multimodal tasks, yet they still hallucinate objects or generate unsafe text. Current hallucination detectors, e.g., single-token linear probing SLP and PTrue, typically analyze only the logit of the first generated token or just its...

MPMA: Preference Manipulation Attack against Model Context Protocol

Model Context Protocol MCP standardizes interface mapping for large language models LLMs to access external data and tools, which revolutionizes the paradigm of tool selection and facilitates the rapid expansion of the LLM agent tool ecosystem. However, as the MCP is increasingly adopted,...

CVE-2025-4701 VITA-MLLM Freeze-Omni utils.py torch.load deserialization

A vulnerability, which was classified as problematic, has been found in VITA-MLLM Freeze-Omni up to 20250421. This issue affects the function torch.load of the file models/utils.py. The manipulation of the argument path leads to deserialization. It is possible to launch the attack on the local ho...

Developer Leaks API Key for Private Tesla, SpaceX LLMs

In AI, as with so many advancing technologies, security often lags innovation. The xAI incident, during which a sensitive API key remained exposed for nearly two months, is a stark reminder of this disconnect. Such oversights not only jeopardize proprietary technologies but also highlight systemi...

S3C2 Summit 2024-09: Industry Secure Software Supply Chain Summit

While providing economic and software development value, software supply chains are only as strong as their weakest link. Over the past several years, there has been an exponential increase in cyberattacks, specifically targeting vulnerable links in critical software supply chains. These attacks...

PT-2025-21290 · Unknown +1 · Vita-Mllm Freeze-Omni +1

Name of the Vulnerable Software and Affected Versions: VITA-MLLM Freeze-Omni up to 20250421 Description: A problematic issue has been found, affecting the torch.load function in the models/utils.py file. The manipulation of the path argument leads to deserialization, allowing an attack to be...

Freeze-Omni 代码问题漏洞

Freeze-Omni is an intelligent low-latency speech dialog model based on Freeze LLM open-sourced by VITA-MLLM. A code issue vulnerability exists in Freeze-Omni 20250421 and earlier versions, which stems from deserialization due to incorrect manipulation of the parameter path in the file...

Dark LLMs: the Growing Threat of Unaligned AI Models

Large Language Models LLMs rapidly reshape modern life, advancing fields from healthcare to education and beyond. However, alongside their remarkable capabilities lies a significant threat: the susceptibility of these models to jailbreaking. The fundamental vulnerability of LLMs to jailbreak...

Cape: Context-Aware Prompt Perturbation Mechanism with Differential Privacy

Large Language Models LLMs have gained significant popularity due to their remarkable capabilities in text understanding and generation. However, despite their widespread deployment in inference services such as ChatGPT, concerns about the potential leakage of sensitive user data have arisen...