kong

Kong : L'ingénieur inverse agentique Orchestration LLM pour la rétro-ingénierie de binaires Qu'est-ce que Kong ? La plupart des tâches suivent une relation linéaire : plus une tâche est difficile, plus elle prend généralement de temps. La rétro-ingénierie et l'analyse binaire est une tâche où la...

inspect_petri

Inspect Petri Bienvenue dans Inspect Petri, un agent d'audit qui permet la surveillance automatisée et l'interaction avec les modèles de langage pour détecter les problèmes d'alignement potentiels, le piratage de récompenses et d'autres comportements préoccupants. Petri vous aide à tester...

oss-fuzz-gen

Платформа для генерации и оценки фаззинг-таргетов Эта платформа генерирует фаззинг-таргеты для реальных проектов на C/C++/Java/Python с помощью различных больших языковых моделей LLM и проводит их бенчмаркинг на платформе OSS-Fuzz. Более подробная информация доступна в статье Фаззинг на основе ИИ...

www-project-top-10-for-large-language-model-applications

www-project-top-10-for-large-language-model-applications Dépôt Web de la Fondation OWASP OWASP Top 10 pour les applications de grands modèles de langage Bienvenue sur le dépôt officiel du OWASP Top 10 pour les applications de grands modèles de langage ! À propos de ce dépôt Ce dépôt contient le...

aegis-audio-defense

AEGIS: Audio Endogenous Guarding via Internal Signals Against Large Audio-Language Model Jailbreaks Submitted to ICASSP 2027. AEGIS keeps the target audio-language model frozen and adds a mid-layer risk gate that conditionally drives late-layer safety adapters, with closed-loop scaling at...

markdown-exfil-tester

valtik-markdown-exfil-tester Testeur en boîte noire pour l'injection indirecte d'invites menant à l'exfiltration de markdown / HTML dans les chatbots basés sur des LLM. Il s'agit de la classe de vulnérabilité qui a touché Microsoft 365 Copilot CVE-2025-32711, alias EchoLeak, ChatGPT correctif...

deleting-the-trace

Атаки с внедрением управляющих токенов на агентов, использующих инструменты Код и зарегистрированные измерения для исследования двух атак на уровне входных данных на агентов на основе языковых моделей, использующих инструменты. Первая атака добавляет короткую строку собственных управляющих токено...

EntropicaPublic

Mises à jour de mars Je dois transférer une partie du travail depuis les dépôts privés. L'exploration de l'utilisation de modèles à très multiples têtes avec des fonctions de mixage et de lecture séparées se poursuit. Actuellement, j'entraîne un modèle Chinese TinyStories sur un modèle équivalent...

batch_jailbreak

Официальный репозиторий для Safety in Batches? Understanding and Mitigating Safety Failures in Batch Prompting Kihyun Kim, Hee-Seon Kim, Wonjun Lee, Changick Kim Korea Advanced Institute of Science and Technology KAIST Новости 2026.09 Наша статья принята на AACL-IJCNLP 2026 Main! 🎉 2026.08 Статья...

birdy-edwards

BIRDY-EDWARDS 2.0 Infiltrate & Expose Автоматизированный инструмент разведки по Facebook на базе ИИ для профилирования целей, анализа сетей и составления отчетов об угрозах. Работает полностью на устройстве через Ollama. SOCMINT-платформа на базе ИИ — 100% локально, без облачной зависимости. Для...

CVE-2026-54559

PocketSphinx is a small speech recognizer. Prior to 5.1.1, the trie language-model loaders in src/lm/ngrammodeltrie.c do not adequately validate boundary conditions in ARPA, DMP, and binary format headers, and the acoustic-model loaders in src/mdef.c and src/util/bio.c use sscanf with unbounded...

CVE-2026-54559 PocketSphinx: Buffer overflows in language and acoustic model loading code

PocketSphinx is a small speech recognizer. Prior to 5.1.1, the trie language-model loaders in src/lm/ngrammodeltrie.c do not adequately validate boundary conditions in ARPA, DMP, and binary format headers, and the acoustic-model loaders in src/mdef.c and src/util/bio.c use sscanf with unbounded...

CVE-2026-54559

PocketSphinx (a lightweight speech recognizer, including its Python bindings) versions 5prealpha through 5.1.0 are affected by buffer overflow vulnerabilities in its model-loading code. The trie language-model loaders in src/lm/ngram_model_trie.c fail to validate boundary conditions in ARPA, DMP,...

CVE-2026-54559 PocketSphinx: Buffer overflows in language and acoustic model loading code

PocketSphinx is a small speech recognizer. Prior to 5.1.1, the trie language-model loaders in src/lm/ngrammodeltrie.c do not adequately validate boundary conditions in ARPA, DMP, and binary format headers, and the acoustic-model loaders in src/mdef.c and src/util/bio.c use sscanf with unbounded...

EEF-CVE-2026-78230 AshAi aggregate tool can read field-policy-protected fields

Summary AshAi exposes Ash read actions to language-model tool calls. The read tool accepts an aggregate result type min, max, sum, avg that builds an ad-hoc Ash.Query.Aggregate over a named field and returns its raw value. Ash field policies redact forbidden fields on returned records replacing...

EUVD-2026-72924

AshAi exposes Ash read actions to language-model tool calls. The read tool accepts an aggregate result type min, max, sum, avg that builds an ad-hoc Ash.Query.Aggregate over a named field and returns its raw value. Ash field policies redact forbidden fields on returned records replacing them with...

Using LLMs to Elicit Security Requirements for Service-Oriented Cyber Ranges

Cyber ranges are complex environments comprising many interacting components and stakeholders with different security concerns. The Service-Oriented Cyber Range SOR is no exception, particularly when it comes to training scenarios targeting critical infrastructure. Security concerns are translate...

PT-2026-81211

U-Boot before 2026.10-rc5 contains a use-after-free vulnerability in the httpc recv cb function within the lwIP wget implementation. When HTTP data storage fails, the callback frees the connection PCB but returns ERR BUF instead of ERR ABRT, causing the TCP input path to access released memory an...

CVE-2026-72742

DSPy 3.3.0b1 contains a file exfiltration vulnerability in the Image and Audio output field adapters that allows attackers with influence over language model outputs to read arbitrary local files by injecting a filesystem path into the url field of a parsed Image or Audio typed output. The...

EUVD-2026-56886

DSPy 3.3.0b1 contains a file exfiltration vulnerability in the Image and Audio output field adapters that allows attackers with influence over language model outputs to read arbitrary local files by injecting a filesystem path into the url field of a parsed Image or Audio typed output. The...