AI Paper Atlas

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

DriveWorld-VLA: Unified Latent-Space World Modeling with Vision–Language–Action for Autonomous Driving

Feiyang Jia, Lin Liu, Ziying Song, Caiyan Jia, Hangjun Ye, Xiaoshuai Hao, Long Chen

ICML 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 0 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 Stage 1 joint VLA and world-model training (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 87.1 -> 87.6Improved
add Stage 2 action controllability fine-tuning (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 87.6 -> 89.5Improved
add Stage 3 future-guided evaluation and refinement (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 89.5 -> 91.3Improved
add Stage 1 training (ablation)DriveWorld-VLA · nuScenes avg CRcollision rate (%): 0.34 -> 0.25Improved
add Stage 2 training (ablation)DriveWorld-VLA · nuScenes avg CRcollision rate (%): 0.25 -> 0.19Improved
add Stage 3 training (ablation)DriveWorld-VLA · nuScenes avg CRcollision rate (%): 0.19 -> 0.16Improved
add Stage 3 for short-horizon planning (ablation)DriveWorld-VLA · nuScenes CR@2s and CR@3scollision rate (%): -0.01 and -0.02No clear effect
use progressive staged training instead of simultaneous Stage 2 and 3 (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 83.6 -> 91.3Improved
unfreeze VLM in Stage 1 (with pre-training) (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 87.6 -> 88.3Improved
add RecogDrive-style pre-training with frozen VLM (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 87.2 -> 87.6Improved
add feature-level supervision from the denoising process (ablation)DriveWorld-VLA · NAVSIMv1PDMS: 87.9 -> 91.3Improved

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