MINERVA：操作策略多小才能解出 LIBERO？
原标题：MINERVA: How Small Can a Manipulation Policy Be and Still Solve LIBERO?
技术动态AI 85
来源：arXiv cs.RO发布时间待核实
arXiv:2609.03715v1 Announce Type: new Abstract: Vision-language-action (VLA) models with billions of parameters now dominate the LIBERO manipulation benchmark, but the model capacity actually required by the benchmark remains unclear. We introduce MINERVA (MINimal Efficient Robotic Vision-Action policy), a family of deliberately compact visuomotor policies designed to measure this task-specific capacity floor.