
recipe-blog-encoding
Hide valuable information where it's least expected

Hide valuable information where it's least expected

Ensemble framework for software vulnerability detection and repair using multiple large language models, with consensus analysis and evaluation tools…

Hybrid machine-learning pipelines for detecting SQL injection in web traffic, combining DistilBERT and BERT-GNN models with adversarial training and…

Implements machine and deep learning methods for indoor UWB jammer localization, including hyperparameter optimization, classification, and…

An automated, high-precision zero-shot evaluation pipeline for OpenAI's CLIP model on CIFAR-10. Features 88.80% accuracy, Safetensors security…

reverse engineering Gemini's SynthID detection

Invisible Hyperlinks in Physical Photographs

Train, evaluate, and explore neural networks with built-in adversarial robustness tools, including PGD attacks, adversarial training, and robust…

Protection against Model Serialization Attacks


A Ghidra plugin for locating object file boundaries.

NotebookLM on steroids — purpose-built for security researchers.

Inference scaling for LLM safety assurance

Detects LLM context-leakage attacks by training lightweight behavior probes on log-probabilities, with vLLM offline/server detection pipelines.

Fully automatic censorship removal for language models

ESPectre - Motion detection system based on Wi-Fi spectre analysis (CSI), with Home Assistant integration.

A unified framework for privacy-preserving data analysis and machine learning

AI / LLM Red Team Field Manual & Consultant’s Handbook