Submission evaluation and evaluation scripts for the Large Language Model Capture-the-Flag (LLM CTF) Competition @ SaTML 2024 of team RSLLM.
Set your API key
export API_KEY="your_api_key"
Script to automate the defense testing process. Goes through all the attacks and evalutes them for a given defense.
python scripts/testing_defense.py
Script to automate the attack competition. Starts an attack and allows for interactive prompting in the competition.
python scripts/competition_attack.py
attacks/sample_attacks.yaml
: contains a list of sample attacksdefenses/gpt.json
: GPT defense (prompt and filters)defenses/llama.json
: LLAMA defense (prompt and filters)
Robin Schmid · GitHub @RobinSchmid7 · Email [email protected]
Takahiro Miki · GitHub @mktk1117 · Email [email protected]
Victor Klemm · GitHub @vklemm · Email [email protected]
Chenhao Li · GitHub @breadli428 · Email [email protected]
Stefan Kraft · GitHub @stekra · Email [email protected]