LLM News

Researchers have introduced RoboGuard, a safety framework that addresses the gap between LLM vulnerabilities and physical robot risks. The system uses a root-of-trust LLM to contextualize safety rules and temporal logic control to prevent harmful robot actions, reducing unsafe plan execution from over 92% to below 3% in tests.

A new research paper introduces methods for predicting when humans will intervene in autonomous web agents by analyzing distinct interaction patterns. The work, which includes a dataset of 400 real-user web navigation trajectories with over 4,200 interleaved human-agent actions, shows that intervention-aware models improved agent usefulness by 26.5% in user studies.