Toward Safe LLM Agents: A Survey of Specification, Verification, and Enforcement: LLM agents increasingly perform irreversible...
Real-world actions, including database updates, API calls, file operations, and autonomous use of tools. However, no existing system provides formally grounded, task-level safety guarantees for the plans these agents generate. Research remains fragmented across specification, verification, and enforcement, limi...
来源简报
Toward Safe LLM Agents: A Survey of Specification, Verification, and Enforcement: LLM agents increasingly perform irreversible real-world actions, including database updates, API calls, file operations, and autonomous use of tools. However, no existing system provides formally grounded, task-level safety guarantees for the plans these agents generate. Research remains fragmented across specification, verification, and enforcement, limi...