📄 中文摘要
强化学习（RL）已成为端到端自动驾驶（AD）的主导范式。然而，RL在复杂场景中面临样本效率低下和缺乏语义可解释性的问题。基础模型，特别是视觉-语言模型（VLMs），能够通过提供丰富的上下文感知知识来缓解这些问题，但其高推理延迟限制了在高频率RL训练循环中的应用。为了解决这一问题，提出了Found-RL平台，旨在高效地利用基础模型增强自动驾驶的强化学习。其核心创新是异步批量推理框架，该框架将重型VLM推理与仿真循环解耦，有效解决延迟瓶颈，从而支持实时学习。
Found-RL：基础模型增强的强化学习用于自动驾驶
📄 English Summary
Found-RL: foundation model-enhanced reinforcement learning for autonomous driving
Reinforcement Learning (RL) has emerged as a dominant paradigm for end-to-end autonomous driving (AD). However, RL suffers from sample inefficiency and a lack of semantic interpretability in complex scenarios. Foundation Models, particularly Vision-Language Models (VLMs), can mitigate these issues by providing rich, context-aware knowledge, yet their high inference latency hinders deployment in high-frequency RL training loops. To address this challenge, Found-RL is proposed as a platform designed to efficiently enhance RL for AD using foundation models. A key innovation is the asynchronous batch inference framework, which decouples heavy VLM reasoning from the simulation loop, effectively resolving latency bottlenecks to support real-time learning.