garak checks if an LLM can be made to fail in a way we don’t want. garak probes for hallucination, data leakage, prompt injection, misinformation, toxicity generation, jailbreaks, and many other weaknesses. If you know nmap or msf / Metasploit Framework, garak does somewhat similar things to them, but for LLMs.
garak focuses on ways of making an LLM or dialog system fail. It combines static, dynamic, and adaptive probes to explore this.
garak‘s a free tool. We love developing it and are always interested in adding functionality to support applications.
currently supports:
garak is a command-line tool. It’s developed in Linux and OSX.
pipJust grab it from PyPI and you should be good to go:
python -m pip install -U garakpipThe standard pip version of garak is updated periodically. To get a fresher version from GitHub, try:
python -m pip install -U git+https://github.com/NVIDIA/garak.git@mainFor more information click here.
The cp command, short for "copy," is the main Linux utility for duplicating files and directories. Whether…
Introduction In digital investigations, images often hold more information than meets the eye. With the…
The cat command short for concatenate, It is a fast and versatile tool for viewing and merging…
What is a Port? A port in networking acts like a gateway that directs data…
The ls command is fundamental for anyone working with Linux. It’s used to display the files and…
The pwd (Print Working Directory) command is essential for navigating the Linux filesystem. It instantly shows your…