AI Paper Atlas

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

Dismantling the Illusion of Vision-Language-Action Models Competence via Explicit Distributional Shifts

Xueyang Zhou, Yangming Xu, Guiyao Tie, Chaoran Hu, Bo Tao, xingwei zhao, Xiang Xiang, Pan Zhou, Lichao Sun, Yongchao Chen

ICML 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
use structured Stair sampling instead of random sampling (ablation)Pi0.5 · LIBERO-Gen Position generalizationsuccess rate (%): 52.3 -> 64.0Improved
use structured Stair sampling instead of random sampling (ablation)Pi0.5 · LIBERO-Gen Task generalizationsuccess rate (%): 14.2 -> 21.2Improved
fine-tune on LIBERO-Gen structured subsetsPi0 · LIBERO-Gen Position (k2)success rate (%): 7.2 -> 13.2Improved
fine-tune on LIBERO-Gen structured subsetsPi0.5 · LIBERO-Gen Position (k2)success rate (%): 49.2 -> 64.0Improved
fine-tune on LIBERO-Gen structured subsetsPi0 · LIBERO-Gen Task (k2)success rate (%): 13.6 -> 13.6No clear effect
add consistency-regularized augmentation (ablation)OpenVLA · LIBERO-Gen Language (k1)accuracy (%): 0.0 -> 98.5Improved
add consistency-regularized augmentation (ablation)OpenVLA · LIBERO-Gen Language (k3, DG)accuracy (%): 39.2 -> 98.6Improved
add consistency-regularized augmentation (ablation)OpenVLA · LIBERO-Gen Object (k2)accuracy (%): 63.8 -> 88.0Improved
add consistency-regularized augmentation (ablation)OpenVLA · LIBERO-Gen Background (k2)accuracy (%): 94.0 -> 98.8Improved
add consistency-regularized augmentation (ablation)OpenVLA · LIBERO-Gen Background (k3, DG)accuracy (%): 99.4 -> 98.8No clear effect
fine-tune on LIBERO-GenOpenVLA · LIBERO-Gen Language (k1)accuracy (%): 0.0 -> 91.4Improved
fine-tune on LIBERO-GenOpenVLA · LIBERO-Gen Background (k1, ID)accuracy (%): 99.6 -> 98.8No 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