Improving Large Vision and Language Models by Learning from a Panel of Peers Paper • 2509.01610 • Published Sep 1 • 2