better_opts_attacks

Posso avere la vostra attenzione? Superare le difese contro le prompt injection basate su fine-tuning con attacchi che sfruttano l'architettura Questo repository contiene il codice per eseguire gli attacchi ASTRA e ASTRA++ che superano SecAlign++, SecAlign e StruQ. Il repository contiene anche...

May I Have Your Attention? Breaking Fine-Tuning Based Prompt Injection Defenses Using Architecture-Aware Attacks

A popular class of defenses against prompt injection attacks on large language models LLMs relies on fine-tuning the model to separate instructions and data, so that the LLM does not follow instructions that might be present with data. There are several academic systems and production-level...