926 matches found
AlphaSteer: Learning Refusal Steering with Principled Null-Space Constraint
As LLMs are increasingly deployed in real-world applications, ensuring their ability to refuse malicious prompts, especially jailbreak attacks, is essential for safe and reliable use. Recently, activation steering has emerged as an effective approach for enhancing LLM safety by adding a refusal...
Privacy Amplification through Synthetic Data: Insights from Linear Regression
Synthetic data inherits the differential privacy guarantees of the model used to generate it. Additionally, synthetic data may benefit from privacy amplification when the generative model is kept hidden. While empirical studies suggest this phenomenon, a rigorous theoretical understanding is stil...
Important: Red Hat Security Advisory: firefox security update
An update for firefox is now available for Red Hat Enterprise Linux 9.2 Extended Update Support. Red Hat Product Security has rated this update as having a security impact of Important. A Common Vulnerability Scoring System CVSS base score, which gives a detailed severity rating, is available for...
firefox: thunderbird: Out-of-bounds access when optimizing linear sums
A flaw was found in Firefox and Thunderbird. The Mozilla Foundation's Security Advisory describes the following issue: An attacker could perform an out-of-bounds read or write on a JavaScript object by confusing array index sizes...
Nearly-Linear Time Private Hypothesis Selection with the Optimal Approximation Factor
Estimating the density of a distribution from its samples is a fundamental problem in statistics. Hypothesis selection addresses the setting where, in addition to a sample set, we are given $n$ candidate distributions -- referred to as hypotheses -- and the goal is to determine which one best...
Robust and Verifiable MPC with Applications to Linear Machine Learning Inference
In this work, we present an efficient secure multi-party computation MPC protocol that provides strong security guarantees in settings with dishonest majority of participants who may behave arbitrarily. Unlike the popular MPC implementation known as SPDZ Crypto '12, which only ensures security wi...
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...
LPASS: Linear Probes As Stepping Stones for Vulnerability Detection Using Compressed LLMs
Large Language Models LLMs are being extensively used for cybersecurity purposes. One of them is the detection of vulnerable codes. For the sake of efficiency and effectiveness, compression and fine-tuning techniques are being developed, respectively. However, they involve spending substantial...
Security update for MozillaFirefox
This update for MozillaFirefox fixes the following issues: Update to Firefox Extended Support Release 128.10.1 ESR. MFSA 2025-37 bsc1243303 CVE-2025-4918: Out-of-bounds access when resolving Promise objects CVE-2025-4919: Out-of-bounds access when optimizing linear sums Patch Instructions: To...
Parallel Kac'S Walk Generates PRU
Ma and Huang recently proved that the PFC construction, introduced by Metger, Poremba, Sinha and Yuen MPSY24, gives an adaptive-secure pseudorandom unitary family PRU. Their proof developed a new path recording technique MH24. In this work, we show that a linear number of sequential repetitions o...
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...
CVE-2024-52426
Improper Neutralization of Input During Web Page Generation 'Cross-site Scripting' vulnerability in Linear Oy Linear linear allows DOM-Based XSS.This issue affects Linear: from n/a through = 2.8.0...
CVE-2024-12496
The Linear plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the plugin's 'linearblockbuycommissions' shortcode in all versions up to, and including, 2.7.12 due to insufficient input sanitization and output escaping on user supplied attributes. This makes it possible for...
CVE-2023-26489
wasmtime is a fast and secure runtime for WebAssembly. In affected versions wasmtime's code generator, Cranelift, has a bug on x8664 targets where address-mode computation mistakenly would calculate a 35-bit effective address instead of WebAssembly's defined 33-bit effective address. This bug mea...
CVE-2022-46381
Certain Linear eMerge E3-Series devices are vulnerable to XSS via the type parameter e.g., to the badging/badgetemplatev0.php component. This affects 0.32-08f, 0.32-07p, 0.32-07e, 0.32-09c, 0.32-09b, 0.32-09a, and 0.32-08e...
Adaptively Secure Distributed Broadcast Encryption with Linear-Size Public Parameters
Distributed broadcast encryption DBE is a variant of broadcast encryption BE that can efficiently transmit a message to a subset of users, in which users independently generate user private keys and user public keys instead of a central trusted authority generating user keys. In this paper, we...
CVE-2022-42710
Nice formerly Nortek Linear eMerge E3-Series 0.32-08f, 0.32-07p, 0.32-07e, 0.32-09c, 0.32-09b, 0.32-09a, and 0.32-08e devices are vulnerable to Stored Cross-Site Scripting XSS...
CVE-2022-31499
Nortek Linear eMerge E3-Series devices before 0.32-08f allow an unauthenticated attacker to inject OS commands via ReaderNo. NOTE: this issue exists because of an incomplete fix for CVE-2019-7256...
CVE-2010-2469
The Linear eMerge 50 and 5000 uses a default password of eMerge for the IEIeMerge account, which makes it easier for remote attackers to obtain Video Recorder data by establishing a session to the device...