Balanced Token Pruning: Accelerating Vision Language Models Beyond Local Optimization

VLM
Re-assign community