AI Paper Atlas

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

Learning to Move Before Learning to Do: Task-Agnostic pretraining for VLAs

Junhao Shi, Siyin Wang, Xiaopeng Yu, Li Ji, Jingjing Gong, Xipeng Qiu

ICML 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 0 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 task-agnostic inverse-dynamics pretraining (TAP-20k) before behavior cloningQwen2.5-VL-3B VLA · SIMPLER WidowXAvg-All success: 23.15% -> 33.32%Improved
add task-agnostic inverse-dynamics pretrainingQwen2.5-VL-3B VLA · SIMPLER WidowXAvg-Partial success: 31.79% -> 45.82%Improved
increase Stage 1 pretraining episodes from 8k to 14k to 20k (ablation)TAP · SIMPLER WidowXAvg-All success: 24.47% -> 30.21% -> 33.32%Improved
scale Stage 1 pretraining steps (20k to 100k) with more Stage 2 finetuning (ablation)TAP · SIMPLER WidowXAvg-All success: stagnates near 18% at 20k steps; exceeds 30% at 100k stepsImproved
extend Stage 2 finetuning without enough Stage 1 pretraining (ablation)TAP · SIMPLER WidowXAvg-All success: stagnates near 18%No clear effect
add task-agnostic pretraining on 30 hours of self-exploration (real world, push task, averaged)WidowX policy with 200 expert demos · Real WidowX push pumpkinsuccess rate: 21% -> 61%Improved
add task-agnostic pretraining (real world, carrot task, averaged)WidowX policy with 200 expert demos · Real WidowX put carrot on platesuccess rate: 9% -> 28%Improved
task-agnostic pretraining under camera viewpoint perturbationWidowX policy · Real WidowX Viewpoint Variationsuccess rate: 0% -> 15% (carrot), 0% -> 25% (pushing)Improved
task-agnostic pretraining in clean standard setup versus large-scale OXE-pretrained NORATAP vs NORA · Real WidowX Standard Setupsuccess rate: TAP 75% vs NORA 85% (push); TAP 40% vs NORA 65% (carrot)Hurt
add task-agnostic pretraining effect on convergence speed during Stage 2 (ablation)TAP · SIMPLER WidowXconvergence speed: baseline saturates around 23%; pretrained exceed 30%No clear effect

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