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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add Stage 1 joint VLA and world-model training (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 87.1 -> 87.6 | Improved |
| add Stage 2 action controllability fine-tuning (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 87.6 -> 89.5 | Improved |
| add Stage 3 future-guided evaluation and refinement (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 89.5 -> 91.3 | Improved |
| add Stage 1 training (ablation) | DriveWorld-VLA · nuScenes avg CR | collision rate (%): 0.34 -> 0.25 | Improved |
| add Stage 2 training (ablation) | DriveWorld-VLA · nuScenes avg CR | collision rate (%): 0.25 -> 0.19 | Improved |
| add Stage 3 training (ablation) | DriveWorld-VLA · nuScenes avg CR | collision rate (%): 0.19 -> 0.16 | Improved |
| add Stage 3 for short-horizon planning (ablation) | DriveWorld-VLA · nuScenes CR@2s and CR@3s | collision rate (%): -0.01 and -0.02 | No clear effect |
| use progressive staged training instead of simultaneous Stage 2 and 3 (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 83.6 -> 91.3 | Improved |
| unfreeze VLM in Stage 1 (with pre-training) (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 87.6 -> 88.3 | Improved |
| add RecogDrive-style pre-training with frozen VLM (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 87.2 -> 87.6 | Improved |
| add feature-level supervision from the denoising process (ablation) | DriveWorld-VLA · NAVSIMv1 | PDMS: 87.9 -> 91.3 | Improved |
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.
- DriveVLA-W0 — DriveVLA-W0: World Models Amplify Data Scaling Law in Autonomous Driving (ICLR 2026) · better on 2 of 2 shared results
- Recogdrive — ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving (ICLR 2026) · better on 1 of 1 shared result
- AutoVLA — AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning (NeurIPS 2025) · better on 1 of 1 shared result
- DrivingGPT — DrivingGPT: Unifying Driving World Modeling and Planning with Multi-modal Autoregressive Transformers (ICCV 2025) · better on 1 of 1 shared result
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
- LaST-VLA: Thinking in Latent Spatio-Temporal Space for Vision-Language-Action in Autonomous Driving (NeurIPS 2026 · similarity 0.92)
- Impromptu VLA: Open Weights and Open Data for Driving Vision-Language-Action Models (NeurIPS 2025 · similarity 0.91)
- AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning (NeurIPS 2025 · similarity 0.91)
- OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation (NeurIPS 2026 · similarity 0.90)
- Less Language, More Latents: Annotation-Efficient VLAs for Driving (NeurIPS 2026 · similarity 0.90)
- FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Driving (NeurIPS 2025 · similarity 0.90)
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.
