AI Paper Atlas

Research map / Video / Video generation

DriveVA: Video Action Models are Zero-Shot Drivers

Mengmeng Liu, Diankun Zhang, Jiuming Liu, Jianfeng Cui, Hongwei Xie, Guang Chen, Hangjun Ye, Michael Yang, Francesco Nex, Hao Cheng

ECCV 2026 · Video › Video generation · approach: World-model video generation · 23 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 video loss (video supervision) (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 71.4 -> 90.9Improved
add video continuation (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 84.6 -> 90.9Improved
add CARLA mix training (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 90.5 -> 90.9Improved
predict video plus action instead of action only (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 47.0 -> 90.9Improved
use bidirectional video-action attention instead of causal mask (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 90.1 -> 90.9Improved
increase sampling steps from 1 to 2 (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 13.2 -> 90.9Improved
increase sampling steps from 2 to 3 (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 90.9 -> 90.9No clear effect
use 8 future frames instead of 4 (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 82.1 -> 90.9Improved
use 12 future frames instead of 8 (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 90.9 -> 86.7Hurt
full fine-tune instead of LoRA on 5B backbone (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 74.9 -> 90.9Improved
pretrained video prior with full fine-tune instead of training from scratch (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 62.9 -> 90.9Improved
scale LoRA backbone from 5B to 14B (ablation)DriveVA · NAVSIM v1 navtest ablationPDMS: 74.9 -> 80.6Improved

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.

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