Paper accepted at RLxF: Reinforcement Learning from World Feedback @ ICML 2026: VLFEEDBACK-EEG: Neural Signals as Implicit Feedback for Vision-Language Model Alignment.