Giskar Red

Giskard Red is a online challenge proposed by Giskard, allowing to explore the limits of large language models.

Hack the LLMs to disrupt their formatting, generate unintended content, or spread misinformation, and secure your spot on the leaderboard.

• https://medium.com/@mehdimerai/an-approach-to-jailbreak-llms-and-bypass-refusals-tested-on-gpt4o-and-gemini-56fca2839221
• https://www.confident-ai.com/blog/how-to-jailbreak-llms-one-step-at-a-time
• https://www.promptfoo.dev/blog/how-to-jailbreak-llms/
• https://www.lakera.ai/blog/jailbreaking-large-language-models-guide
• https://arxiv.org/pdf/2307.02483