面向规划的 token 压缩：长上下文驾驶模型提速 Planning-aligned Token Compression for Long-Context Autonomous Driving

精读摘要

一体化视觉-动作模型是自动驾驶的新兴范式，但编码长时序上下文时产生的 token 序列会迅速超出实时计算预算。线性注意力、外部记忆等方案让上下文变轻，但 token 压缩最契合架构——它无需改动骨干网络。现有压缩却采用时间衰减、解耦融合等规则式启发法；这项研究提出与规划对齐的 token 压缩，让压缩服务于驾驶规划本身，为长上下文自动驾驶模型实时运行提供新方案。 Monolithic vision-action models produce token sequences that exceed real-time computational budgets when encoding extended temporal context. Token compression is the most architecture-compatible solution, requiring no backbone modifications, but existing methods use rule-based heuristics like temporal decay. This work proposes planning-aligned token compression so that compression serves driving planning itself.