RoboTrustBench: Benchmarking the Trustworthiness of Video World Models for Robotic Manipulation
技术动态
来源：arXiv cs.RO发布时间待核实
arXiv:2606.01600v2 Announce Type: replace-cross Abstract: Video world models are increasingly used in robotic manipulation, yet existing benchmarks mostly evaluate them under valid, feasible, and safe instructions. We introduce RoboTrustBench, a benchmark for evaluating the trustworthiness of video world models under four scenarios: Normal, Constraint-Sensitive, Counterfactual, and Adversarial.