1 dépôt
Awesome GitHub RepositoriesAdversarial Suffix Optimization
Applying gradient descent to optimize string sequences that trigger prohibited model responses.
Distinct from Gradient Descent Algorithms: Focuses on optimizing input tokens for adversarial attack rather than updating model parameters for training.
Explore 1 awesome GitHub repository matching artificial intelligence & ml · Adversarial Suffix Optimization. Refine with filters or upvote what's useful.
Awesome Adversarial Suffix Optimization GitHub Repositories
Trouvez les meilleurs dépôts grâce à l'IA.Nous recherchons les dépôts les plus pertinents grâce à l'IA.
- llm-attacks/llm-attacks
llm-attacks/llm-attacks
4,509Voir sur GitHubThis repository provides tools and methodologies for studying adversarial attacks on large language models. It focuses on understanding how carefully crafted inputs can manipulate or bypass the safety mechanisms of LLMs, enabling researchers to probe model vulnerabilities and improve their robustness. The project covers techniques for generating adversarial prompts, evaluating model responses under attack conditions, and analyzing the effectiveness of different attack strategies.
Uses coordinate gradient descent to iteratively refine token sequences that maximize adversarial response probability.
Python
