AgentWatcher

AgentWatcher AgentWatcher es una defensa basada en detección contra la inyección indirecta de prompts en agentes LLM. Primero ejecuta atribución causal de contexto sobre el contexto no confiable para encontrar los contextos más influyentes, y luego aplica un LLM monitor que clasifica esos context...

CVE-2022-0847-eBPF

CVE-2022-0847-eBPF Un programa eBPF para detectar y defender ataques en CVE-2022-0847. En el espacio del kernel, el programa detecta las llamadas al sistema splice ; si dicha llamada contiene un pipe como fd de salida y los flags del buffer del pipe están establecidos en PIPEBUFFLAGCANMERGE ,...

detection-defense-library

Detection Defense Library One defense-engineering library, consolidated. Attack classes, the patterns that detect them, the committed Sigma rules, and the Windows trust-boundary model, all in a single repo. This repo folds in five former sibling repositories: Former repo| Folded into| What it hol...

detection-defense-library — Updated!

Detection Defense Library One defense-engineering library, consolidated. Attack classes, the patterns that detect them, the committed Sigma rules, and the Windows trust-boundary model, all in a single repo. This repo folds in five former sibling repositories: Former repo| Folded into| What it hol...

A Cryptographic Perspective on Mitigation Vs. Detection in Machine Learning

In this paper, we initiate a cryptographically inspired theoretical study of detection versus mitigation of adversarial inputs produced by attackers of Machine Learning algorithms during inference time. We formally define defense by detection DbD and defense by mitigation DbM. Our definitions com...