AutoDojo

Emerging

1papers using it

2026first seen

AutoDojo is an adaptive benchmark that evaluates the robustness of defenses against indirect prompt injection (IPI) attacks in LLM-powered agents by optimizing IPI strategies against specific defenses.

🔎 Find this dataset

Papers using AutoDojo (1)

AutoDojo: Adaptive Attacks Expose Superficial Defenses and User-Underspecification Limits in LLM Agents2026