Abstract page for arXiv paper 2611.08711: RL-ATTACK: A Generalized Black-Box Adversarial Attack on Language Models via Reinforcement Learning