Quincunx33/Ai-jailbreak
A collection of jailbreak prompts and adversarial techniques targeting local and frontier AI models, framed for red-teaming and AI safety research.
Collecting fresh signals — velocity needs a few days of history.
collecting data…
star history
The repository catalogs prompt-based exploit techniques for LLMs, including role confusion, multi-shot dissociation, and logical inversion, and notes the model flaws each technique targets. Entries reference OWASP LLM risk categories and cover models such as Qwen, Gemma, Llama, Kimi, GPT, Gemini, and Grok. The material is presented for red-teaming verification and defensive AI alignment research.
Frequently asked
- What is Quincunx33/Ai-jailbreak?
- A collection of jailbreak prompts and adversarial techniques targeting local and frontier AI models, framed for red-teaming and AI safety research.
- Is Ai-jailbreak open source?
- Yes — Quincunx33/Ai-jailbreak is an open-source project tracked on heatdrop.
- What language is Ai-jailbreak written in?
- Quincunx33/Ai-jailbreak is primarily written in JavaScript.
- How popular is Ai-jailbreak?
- Quincunx33/Ai-jailbreak has 502 stars on GitHub.
- Where can I find Ai-jailbreak?
- Quincunx33/Ai-jailbreak is on GitHub at https://github.com/Quincunx33/Ai-jailbreak.