Announcement_13
New! New preprint out on Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents — project page at unsafer-in-many-turns.github.io, code at CHATS-lab/ToolShield.
New! New preprint out on Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents — project page at unsafer-in-many-turns.github.io, code at CHATS-lab/ToolShield.