Secure Speculative Decoding for Large Language Models

Speculative decoding accelerates inference for a large language model LLM, referred to as the target model, by first using a smaller model, referred to as the draft model, to generate candidate tokens and then verifying them with the target model for acceptance or rejection. Prior studies primari...

The Model Plants the Trigger: Answer-Side Backdoor Attacks in Multi-Turn Large Language Models

Safety alignment in Large Language Models LLMs remains vulnerable to backdoor attacks. Existing LLM backdoors are almost all input-centric: activation depends on explicit trigger patterns in the user input, so modern guardrails are built to sanitize the input space. We challenge this assumption...

Linux Distros Unpatched Vulnerability : CVE-2026-104711

The Linux/Unix host has one or more packages installed that are impacted by a vulnerability without a vendor supplied patch available. - Improper neutralization of special elements used in an expression language statement 'Expression Language Injection' vulnerability in Apache Struts. If the...

CVE-2018-6574

CVE-2018-6574 CVE-2018-6574 go get RCE...

impostor

Impostor Servidor SMB1 para interceptar hashes NTLMv2, escrito en C. El objetivo del proyecto es implementar un servidor SMB1 como una máquina de estados que acepte y responda correctamente a las solicitudes del cliente para engañarlo, llevándolo a completar una autenticación NTLM y así revelar s...

Security Bulletin: Due to the use of Spring Expression Language (SpEL), IBM Sterling Control Center is vulnerable to Improper Control of Dynamically-Managed Code Resources (CVE-2026-59283)

Summary Spring MVC application evaluate Spring Expression Language SpEL which used by IBM Sterling Control Center as part of dependency in spring-expression-6.2.12.jar Vulnerability Details CVEID:CVE-2026-59283 DESCRIPTION: Applications that evaluate Spring Expression Language SpEL expressions...

Security Bulletin: Due to the use of Spring Expression Language (SpEL), IBM Sterling Control Center is vulnerable to Uncontrolled Resource Consumption (CVE-2026-47886)

Summary Spring Expression Language SpEL is used by IBM Sterling Control Center as part of dependency in spring-expression-6.2.12.jar. Vulnerability Details CVEID:CVE-2026-47886 DESCRIPTION: Applications that evaluate user-supplied Spring Expression Language SpEL expressions may be vulnerable to a...

CVE-2026-104711

Improper neutralization of special elements used in an expression language statement 'Expression Language Injection' vulnerability in Apache Struts. If the application is configured to use the legacy RESTful action mapper, a crafted request can inject an OGNL expression that may lead to remote co...

EUVD-2026-92622

Improper neutralization of special elements used in an expression language statement 'Expression Language Injection' vulnerability in Apache Struts. If the application is configured to use the legacy RESTful action mapper, a crafted request can inject an OGNL expression that may lead to remote co...

CVE-2023-4966

CVE-2023-4966 CitrixBleed Programa para explotar aplicaciones Citrix con un rango de direcciones IP Así es como lo compilas para Windows, macOS o Linux GOOS=windows GOARCH=amd64 go build -o citrix citrix.go GOOS=darwin GOARCH=amd64 go build -o citrix.exe citrix.go GOOS=linux GOARCH=amd64 go build...

CVE-2020-13942byVulHub

CVE-2020-13942byVulHub Apache Unomi Inyección de Lenguaje de Expresión RCE CVE-2020-13942 Descripción Esta vulnerabilidad permite la ejecución de comandos arbitrarios en sistemas que ejecutan Apache Unomi versión 1.5.1, explotando la inyección del lenguaje de expresión MVEL. Página del proyecto...

Next.js-Exploit-Tool

Next.js-Exploit-Tool Herramienta integral de explotación con interfaz gráfica para Next.js. Introducción Una herramienta independiente de evaluación de seguridad para CVE-2025-55182 , desarrollada en Go, que ofrece una interfaz gráfica GUI para realizar rápidamente detección y verificación de...

CVE-2026-11105-Stack-Buffer-Overflow-in-Custom-Base64-Decoder

CVE-2026-11105 – Desbordamiento de búfer en la pila en el decodificador Base64 personalizado Código del programa C // base64vuln.c - Vulnerable Base64 decoder include include include int base64decodeconst char in, sizet inlen, char out, sizet outlen static const char table =...

lua_CVE-2020-24370_AfterPatch

Lua 5.3 readme blockquote, .display border: solid a0a0a0 2px ; border-radius: 8px ; padding: 1em ; margin: 0px ; .display word-spacing: 0.25em ; dl.display dd padding-bottom: 0.2em ; tt, kbd, code font-size: 12pt ; Bienvenido a Lua 5.3 acerca de · instalación · cambios · licencia · manual de...

Astra Linux – Vulnerability in Chromium

Before version 148.0.7778.216, using free in Google Chrome’s views allowed a remote attacker who convinced a user to perform certain UI gestures to execute arbitrary code through a crafted HTML page. Chromium security severity: High...

Astra Linux – Vulnerability in Golang-1.19, Golang-1.23

A discrepancy in how Go and C/C++ comments are parsed allowed for code to be smuggled into the resulting cgo binary...

Astra Linux – Vulnerability in Chromium

Uninitialized use of ANGLE in Google Chrome prior to 150.0.7871.46 allowed a remote attacker who had compromised the renderer process to potentially perform a sandbox escape through a crafted HTML page. Chromium security severity: High...

Astra Linux – Vulnerability in Erlang

Erlang/OTP is a set of libraries for the Erlang programming language. Prior to versions OTP-27.3.3, OTP-26.2.5.11, and OTP-25.3.2.20, a SSH server might allow an attacker to perform unauthenticated remote code execution RCE. By exploiting a flaw in the SSH protocol’s message handling, a malicious...

Astra Linux – Vulnerability in Chromium

Inappropriate implementation in the Password Manager in Google Chrome prior to 149.0.7827.53 allowed a remote attacker to leak cross-origin data through a crafted HTML page. Chromium security severity: Medium...

Adaptive_Greedy_Local_Search

Adaptive Greedy Local Search AGLS Secuestro de Prompts con Preservación Semántica: Un Ataque Adversarial de Caja Negra sobre la Optimización Automática de Prompts ICME 2026 Resumen: Los Modelos de Lenguaje de Gran Tamaño LLMs están cada vez más equipados con módulos automáticos de optimización de...