Sattyam Jain

Provael

Sattyam Jain (1 project)

Adversarial red-teaming and assurance for robot vision-language-action policies. Runs safety scenarios against VLA models, reports measured failure rates with the benign control alongside, and main...