AI Paper Atlas

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

Unified Vision-Language-Action Model

Yuqi Wang, Xinghang Li, Wenxuan Wang, Junbo Zhang, Yingyan Li, Yuntao Chen, Xinlong Wang, Zhaoxiang Zhang

ICLR 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 134 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 world model post-training on videos (ablation)UniVLA · LIBEROsuccess rate: 48.5 -> 94.2Improved
add world model post-training on videos (ablation)UniVLA · CALVINaverage length: 1.46 -> 4.61Improved
add world model post-training on videos (ablation)UniVLA · SimplerEnv-WidowXsuccess rate: 0.0 -> 64.6Improved
add world model post-training on videos (ablation)UniVLA · LIBERO-Longsuccess rate: 17.4 -> 89.2Improved
post-train with action prediction only (ablation)UniVLA · LIBEROsuccess rate: 48.5 -> 43.9Hurt
post-train with video prediction instead of world model (no text) (ablation)UniVLA · LIBEROsuccess rate: 94.2 -> 78.9Hurt
post-train with text-to-image instead of world model (ablation)UniVLA · CALVINaverage length: 4.61 -> 3.79Hurt
add post-training with 10% fine-tuning data (ablation)UniVLA · CALVINaverage length: 0.15 -> 3.19Improved
add post-training for faster convergence at 2k iterations (ablation)UniVLA · CALVINaverage length: 0.37 -> 4.21Improved
add visual prediction loss without post-training (ablation)UniVLA · LIBEROsuccess rate: 48.5 -> 88.7Improved
add history window of 10 with 1+1 observations (ablation)UniVLA · CALVINaverage length: 4.26 -> 4.61Improved
extend history from 1+1 to 1+2 observations (ablation)UniVLA · CALVINaverage length: 4.61 -> 4.43Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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