NVIDIA/garak

scans LLMs for vulnerabilities like prompt injection, hallucination, and data leaks

View on index · View in 3D Map

Related repos

• Extracts document content for generative applications
• interactive NVIDIA GPU process viewer
• NVIDIA Isaac GR00T lets robots use language and images to do tasks in different environments
• toolkit for building and optimizing AI agent workflows
• CUDA C++ templates for matrix multiplication and linear algebra
• lets LLMs automate reverse engineering in Ghidra
• lets you unlock NVENC session limits and NvFBC on Nvidia consumer GPUs
• Runs local LLMs in a browser
• interactive GPU monitoring tool for Linux systems