Saved in:
Bibliographic Details
Main Authors: Qu, Yansong, Sheng, Zihao, Huang, Zilin, Chen, Jiancong, Luo, Yuhao, Wang, Tianyi, Feng, Yiheng, Labi, Samuel, Chen, Sikai
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.10458
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912896727908352
author Qu, Yansong
Sheng, Zihao
Huang, Zilin
Chen, Jiancong
Luo, Yuhao
Wang, Tianyi
Feng, Yiheng
Labi, Samuel
Chen, Sikai
author_facet Qu, Yansong
Sheng, Zihao
Huang, Zilin
Chen, Jiancong
Luo, Yuhao
Wang, Tianyi
Feng, Yiheng
Labi, Samuel
Chen, Sikai
contents 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 this by offering rich, context-aware knowledge, yet their high inference latency hinders deployment in high-frequency RL training loops. To bridge this gap, we present Found-RL, a platform tailored to efficiently enhance RL for AD using foundation models. A core 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. We introduce diverse supervision mechanisms: Value-Margin Regularization (VMR) and Advantage-Weighted Action Guidance (AWAG) to effectively distill expert-like VLM action suggestions into the RL policy. Additionally, we adopt high-throughput CLIP for dense reward shaping. We address CLIP's dynamic blindness via Conditional Contrastive Action Alignment, which conditions prompts on discretized speed/command and yields a normalized, margin-based bonus from context-specific action-anchor scoring. Found-RL provides an end-to-end pipeline for fine-tuned VLM integration and shows that a lightweight RL model can achieve near-VLM performance compared with billion-parameter VLMs while sustaining real-time inference (approx. 500 FPS). Code, data, and models will be publicly available at https://github.com/ys-qu/found-rl.
format Preprint
id arxiv_https___arxiv_org_abs_2602_10458
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Found-RL: foundation model-enhanced reinforcement learning for autonomous driving
Qu, Yansong
Sheng, Zihao
Huang, Zilin
Chen, Jiancong
Luo, Yuhao
Wang, Tianyi
Feng, Yiheng
Labi, Samuel
Chen, Sikai
Artificial Intelligence
Machine Learning
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 this by offering rich, context-aware knowledge, yet their high inference latency hinders deployment in high-frequency RL training loops. To bridge this gap, we present Found-RL, a platform tailored to efficiently enhance RL for AD using foundation models. A core 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. We introduce diverse supervision mechanisms: Value-Margin Regularization (VMR) and Advantage-Weighted Action Guidance (AWAG) to effectively distill expert-like VLM action suggestions into the RL policy. Additionally, we adopt high-throughput CLIP for dense reward shaping. We address CLIP's dynamic blindness via Conditional Contrastive Action Alignment, which conditions prompts on discretized speed/command and yields a normalized, margin-based bonus from context-specific action-anchor scoring. Found-RL provides an end-to-end pipeline for fine-tuned VLM integration and shows that a lightweight RL model can achieve near-VLM performance compared with billion-parameter VLMs while sustaining real-time inference (approx. 500 FPS). Code, data, and models will be publicly available at https://github.com/ys-qu/found-rl.
title Found-RL: foundation model-enhanced reinforcement learning for autonomous driving
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2602.10458