外观
论文arXiv 2610.1533717 天前
CoLA：通过学习适配器实现多 VLA 协调
原文链接：CoLA: Multi-VLA Coordination via Learned Adapters（arXiv 2610.15337 · 发布日期：2026-10-11）
CoLA: Multi-VLA Coordination via Learned Adapters
CoLA冻结各智能体VLA骨干，学习消息通道和扩散动作头，在三类多臂交接任务中验证显式通信何时必要。
核心要点
- 研究多VLA智能体共享工作空间时，何时需要显式消息通信而非仅靠相互观察。
- CoLA在冻结VLA骨干上训练轻量消息通道和逐臂扩散动作头，并支持完全去中心化执行。
- 实验显示信息私有时通道才关键：遮挡托盘交接成功率76.0%，零消息训练为1.3%。
- 在共享顶置相机下移除消息影响不明显；仅腕视角二臂任务下降，且架构可跨VLA骨干泛化。
验证范围与局限：验证范围为摘要中的遮挡托盘、二臂和三臂交接任务；摘要未说明仿真或真机。
英文摘要 (Abstract)
Vision-Language-Action (VLA) models are typically deployed as isolated single-agent policies. When several such agents share a workspace, they can coordinate implicitly, by observing one another, or explicitly, by exchanging messages, but it remains unclear when explicit communication becomes necessary. We introduce CoLA (Coordination via Learned Adapters), which keeps each agent's VLA backbone frozen and trains a lightweight message channel and a per-arm diffusion action head on top, with fully decentralised execution. Across three tasks that differ in how much each agent can observe, the channel matters only when information is private. On an occluded tray handover, where the box's colour marker is visible to one agent only, CoLA reaches 76.0% success versus 1.3% with messages zeroed throughout training. On two-arm and three-arm handover tasks, removing messages has no measurable effect with a shared overhead camera (90.6% vs. 88.3% at two arms; 63.5% vs. 60.0% at three), but lowers success on the two-arm task's wrist-only variant (74.8% vs. 64.3%). This architecture generalises across different VLA backbones as well. With a frozen $π_{0.5}$ backbone on the same occluded tray task, CoLA matches the overall success rate of a LoRA-fine-tuned baseline (37.3% vs. 34.7%) using a fraction of the trainable parameters. When CoLA places the box, it selects the correct tray 71.8% of the time, whereas the independently fine-tuned baseline, whose placing arm never sees the marker, selects it at chance level (34.7% against 33.3%).
作者：Ahaskar Kashyap, Jiawen Zong, Boyu Ma, Chuhao Zhou, Jingliang Li, Stefano V. Albrecht, Jianfei Yang
来源
本页事实对应的公开页面，以原文为准。