AI Paper Atlas

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

Vision-Language-Action Instruction Tuning: From Understanding to Manipulation

Shuai Yang, Hao Li, Bin Wang, Yilun Chen, Yang Tian, Tai Wang, Hanqing Wang, Feng Zhao, Yiyi Liao, Jiangmiao Pang

ICLR 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models

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 QA and captioning data to VLA-IT (ablation)InstructVLA-Generalist · SimplerEnv-Instructaverage success rate: 41.7 -> 46.2Improved
enable explicit textual thinking before acting (ablation)InstructVLA-Generalist · SimplerEnv-Instructsuccess rate: 36.1% gain over direct executionImproved
add language motion supervision (ablation)InstructVLA · SimplerEnvoverall success rate: 9.3% improvementImproved
add DINOv2 ViT encoder to the action expert (ablation)InstructVLA action expert · SimplerEnvperformance: removal gives 50.0% dropImproved
add FiLM to the ViT encoder (ablation)InstructVLA action expert · SimplerEnvperformance: 15.3% further improvementImproved
use MoE adaptation and two-stage training instead of full finetuning or autoregressive co-training (ablation)InstructVLA · SimplerEnvsuccess rate: 12.5% improvement over MagmaImproved
remove MoE design while keeping training paradigm (ablation)InstructVLA · SimplerEnvmanipulation and multimodalMixed
freeze action expert during VLA-IT instead of joint finetuning (ablation)InstructVLA · SimplerEnv-Instructsuccess rateNo clear effect
add VLA-IT data on both Bridge and Fractal vs Bridge only (ablation)InstructVLA · SimplerEnv-Instructoverall success rate: 15.6 -> 21.7 -> 46.0Improved
finetune OpenVLA with VLA-IT data on top of multimodal data (ablation)OpenVLA · SimplerEnv-Instruct situated reasoningsuccess rate: 19.5 -> 17.4No clear effect
add GPT-4o as system-2 instruction rewriter to OpenVLA (ablation)OpenVLA + VL · SimplerEnv-Instructaverage success rate: 23.9 -> 35.6Improved
increase number of latent action tokens (16 to 128) (ablation)InstructVLA · SimplerEnvrobot performanceMixed

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