Research map / Vision-language and multimodal / Robot action and navigation
FastDriveVLA: Efficient End-to-End Driving via Plug-and-Play Reconstruction-based Token Pruning
Jiajun Cao, Qizhe Zhang, Peidong Jia, Xuhui Zhao, Bo Lan, Xiaoan Zhang, , Lizhuo, Xiaobao Wei, Sixiang Chen, Liyun Li, Xianming Liu, Ming Lu and 2 more
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 adversarial foreground-background reconstruction (AFBR) to pixel reconstruction (ablation) | ReconPruner · nuScenes open-loop planning (50% tokens pruned, ablation) | Collision avg (%): 0.29 -> 0.25 | Improved |
| add pixel reconstruction (vs foreground mask prediction) to AFBR (ablation) | ReconPruner · nuScenes open-loop planning (50% tokens pruned, ablation) | L2 avg (cm): 32.19 -> 32.10 | Improved |
| use ReconPruner scores instead of ground-truth foreground mask + text attention (ablation) | token pruning · nuScenes open-loop planning (50% tokens pruned, ablation) | Collision avg (%): 0.26 -> 0.25 | Improved |
| restrict text-attention pruning to tokens within ground-truth foreground mask (ablation) | text-attention pruning · nuScenes open-loop planning (50% tokens pruned, ablation) | L2 avg (cm): 32.35 -> 32.16 | Improved |
| use FastDriveVLA instead of unpruned model at 25% pruning | Impromptu-VLA · nuScenes open-loop planning (25% tokens pruned) | L2 avg (cm): 31.83 -> 31.80 | Mixed |
| use FastDriveVLA instead of VisPruner/FastV/DivPrune at 75% pruning | Impromptu-VLA · nuScenes open-loop planning (75% tokens pruned) | L2 avg (cm): 32.77 -> 32.64 | Improved |
| increase pruning ratio from 25% to 50% (collision metric) | token pruning methods · nuScenes open-loop planning | Collision avg (%): FastDriveVLA 0.26 -> 0.25 | No clear effect |
| prune 3249 to 812 visual tokens with FastDriveVLA | Impromptu-VLA · Efficiency | FLOPs (T): 38.2 -> 5.1 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DivPrune — DivPrune: Diversity-based Visual Token Pruning for Large Multimodal Models (CVPR 2025) · better on 6 of 6 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
- Prune2Drive: A Plug-and-Play Framework for Accelerating Vision-Language Models in Autonomous Driving (CVPR 2026 · similarity 0.94)
- Each Complexity Deserves a Pruning Policy (NeurIPS 2025 · similarity 0.91)
- Impromptu VLA: Open Weights and Open Data for Driving Vision-Language-Action Models (NeurIPS 2025 · similarity 0.90)
- FIVE-VLA: Fast and EffectIVE Closed-Loop Autonomous Driving with Recurrent Action Memory (NeurIPS 2026 · similarity 0.90)
- Less Language, More Latents: Annotation-Efficient VLAs for Driving (NeurIPS 2026 · similarity 0.90)
- When and What to Prune? Stage-Aware Visual Token Pruning for Efficient VLA (NeurIPS 2026 · similarity 0.89)
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.
