ToolHazard: Scaling Adversarial Environments for Security Evaluation and Alignment of LLM-based Agents

Part of

Adaptive Interfaces And AI Benchmarking