liu00222/Open-Prompt-Injection
Open-source toolkit and benchmark for prompt injection attacks and defenses on LLMs.
Collecting fresh signals — velocity needs a few days of history.
collecting data…
star history
The repository provides a framework for implementing, evaluating, and extending prompt injection attacks and defenses against large language models. It supports multiple model providers including PaLM2, Llama, and OpenAI GPT, and includes evaluation scenarios combining target tasks with injected tasks. It is intended as a research benchmark for studying LLM security and privacy.
Frequently asked
- What is liu00222/Open-Prompt-Injection?
- Open-source toolkit and benchmark for prompt injection attacks and defenses on LLMs.
- Is Open-Prompt-Injection open source?
- Yes — liu00222/Open-Prompt-Injection is open source, released under the MIT license.
- What language is Open-Prompt-Injection written in?
- liu00222/Open-Prompt-Injection is primarily written in Python.
- How popular is Open-Prompt-Injection?
- liu00222/Open-Prompt-Injection has 500 stars on GitHub.
- Where can I find Open-Prompt-Injection?
- liu00222/Open-Prompt-Injection is on GitHub at https://github.com/liu00222/Open-Prompt-Injection.