ATBench

ATBench is a benchmark used to evaluate the performance of attack detection frameworks on tool-call traffic from LLM agents, focusing on the effectiveness of various architectures and features in classifying sessions as benign or attacked.

Papers using ATBench (6)

AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security2026Ask Now, Use Later: Benchmarking the Proactivity Gap in Long-Lived LLM Agents2026Autonomous computational catalysis through an agentic research system2026 · 1 citesContent-Aware Attack Detection in LLM Agent Tool-Call Traffic: An Empirical Study of Features, Architectures, and Evaluation Protocols2026RAT: RunAnyThing via Fully Automated Environment Configuration2026Aptbench: Benchmarking Agentic Potential Of Base Llms During Pre-training2025