DARWIN is a self-evolving LLM jailbreak framework that grows a reusable strategy pool through external extraction, sandbox filtering, history-memory retrieval, Markov-based strategy selection, and reflective/genetic evolution.
Python Updated 7 months ago 1 commits 8 stars
surge-bot|Mirror: ZJU-LLM-Safety/DARWIN (⭐8)
7 months agoE2B SandboxFirecracker microVMpython
No executable code files found in this repository.
About
DARWIN is a self-evolving LLM jailbreak framework that grows a reusable strategy pool through external extraction, sandbox filtering, history-memory retrieval, Markov-based strategy selection, and reflective/genetic evolution.
Readme
8 stars