AI45Lab
AI45Lab builds research frameworks and tools for AI safety and autonomous agent evaluation. The lab develops AgentDoG, a diagnostic guardrail framework for trajectory-level risk assessment in agentic systems. Their work focuses on fine-grained safety taxonomy, benchmark datasets, and explainability for tool-using AI agents.
AI Tools by AI45Lab
(1)
AgentDoG
AI Agent Safety Guardrail Framework
Discussions
No discussions yet
Be the first to start a discussion about AI45Lab
AI Topics
3
AI45Lab focuses on these topics:
Know more about AI45Lab? Start a discussion to share what you know.