Llama

The 'Llama' dataset/benchmark is used to evaluate the effectiveness of hybrid approaches in exploiting vulnerabilities of Large Language Models (LLMs) through token-level and prompt-level jailbreak strategies.

Papers using Llama (2)

Blind Spots in the Guard: How Domain-Camouflaged Injection Attacks Evade Detection in Multi-Agent LLM Systems2026Advancing Jailbreak Strategies: A Hybrid Approach To Exploiting LLM Vulnerabilities And Bypassing Modern Defenses2025