AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

MindDrive: A Vision-Language-Action Model for Autonomous Driving via Online Reinforcement Learning

haoyu fu, Diankun Zhang, Zongchuang Zhao, Jianfeng Cui, Hongwei Xie, Bing Wang, Guang Chen, Hangjun Ye, Dingkang Liang, Xiang Bai

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 26 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add online RL after imitation learning (ablation)MindDrive (0.5B) · Bench2DriveDriving Score / Success Rate: 75.85 -> 78.04 DS; 49.30 -> 55.09 SRImproved
add online RL after imitation learning (ablation)MindDrive-L (3B) · Bench2DriveDriving Score / Success Rate: 75.78 -> 80.59 DS; 50.71 -> 58.26 SRImproved
add collision penalty (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 49.30 -> 50.70Improved
add traffic light penalty on top of collision (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 50.70 -> 51.47Improved
add route deviation penalty (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 51.47 -> 51.85Improved
add stop sign penalty (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 51.85 -> 55.09Improved
use KL regularization instead of vanilla PPO (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 46.73 -> 55.09Improved
use KL regularization instead of entropy regularization (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 49.24 -> 55.09Improved
use VLM meta-actions instead of navigation commands (IL) (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 41.59 -> 49.30Improved
set GAE lambda to 1.00 (Monte Carlo) instead of 0.90 (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 48.38 -> 55.09Improved
set GAE lambda to 1.00 instead of 0.95 (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 49.03 -> 55.09Improved
train online RL for 2 epochs instead of 1 (ablation)MindDrive (0.5B) · Bench2DriveSuccess Rate (%): 48.61 -> 55.09Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy