ActGuard
binzhwang/actguardPre-execution action-auditing defense that detects and masks indirect prompt injection in tool-using LLM agents using embedding retrieval and…
ai-securitydefensive-toolsmachine-learning+2
binzhwang/actguardPre-execution action-auditing defense that detects and masks indirect prompt injection in tool-using LLM agents using embedding retrieval and…
off-by-1-labs/flawedFix-Like Artifacts With Embedded Defects
c2a2-at-florida-atlantic-university/gr-physecGNU Radio module for physical layer security using deep learning channel fingerprinting, feature quantization, and SHA3-512 key generation from SDR…
xiaoyan-lisa/defenses-for-tool-integrated-llmResearch code and experiments for defending tool-integrated LLM agents against adversarial attacks, extending Agent Security Bench with new defense…
siyuanli00/intactResearch code reproducing multi-turn LLM jailbreak experiments (FITD, MRCJ, ActorAttack, X-Teaming) from the SoK intent-oriented systematization…
zdcslab/rubrics-as-an-attack-surfaceResearch code for Rubric-Induced Preference Drift (RIPD): evolutionary rubric search, benchmark-preserving selection, and DPO policy misalignment…
9xeb/context-segmentationMulti-agent automated context management for long horizon tasks in local AI Agents
svdwi/blueboxOpen-source threat intelligence platform for malware and observable analysis. Enriches IPs, domains, URLs, and hashes with external sources, performs…
davidc1212/crawRobust audio watermarking framework embedding binary messages into magnitude spectrograms, with differentiable attack simulation, Q-Former pooling,…
atsukisato/pgm-attackResearch code for poisoning attacks on the PGM-index, demonstrating how to craft adversarial data to degrade learned index performance.
mlily2024/final-project-sql-injection-pipelineHybrid machine-learning pipelines for detecting SQL injection in web traffic, combining DistilBERT and BERT-GNN models with adversarial training and…
pwnhyo/t-mapTrajectory-aware evolutionary search framework for red-teaming LLM agents over MCP servers, generating adversarial prompts to map vulnerability…
qkrrkd90-source/weavemarkReference implementation of a multi-bit LLM watermarking scheme using coded payload spreading, unbiased reweighting, and soft-decision ECC decoding…
cucu220123/safety-awarenessResearch code for extracting and training safety-awareness directions in multimodal LLMs to improve refusal behavior while limiting benign-task drift.
chanwoochoi316/kidnapragBlack-box attack framework that hijacks reasoning in agentic retrieval-augmented generation systems by injecting poisoned documents, with support for…
shuangliangx/learning-to-detectDetects unknown jailbreak attacks in large vision-language models using hidden state analysis and autoencoders, with training and evaluation…
binzhwang/tnc-defenseResearch code for detecting and detoxifying backdoors in text-to-image diffusion models, with pipelines for Stable Diffusion v1.4, v1.5, XL, and 3…
jianshuod/ipi-exposure-signalResearch pipeline for detecting latent indirect prompt-injection exposure signals in agentic LLMs via hidden-state probing, including trace…