AI Paper Atlas

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

LIBERO-Plus: A Progressive Robustness Benchmark for Visual-Language-Action Models

Senyu Fei, Siyin Wang, Junhao Shi, Zihao Dai, Jikun Cai, Pengfang Qian, Li Ji, Xinzhe He, Shiduo Zhang, Zhaoye Fei, Jinlan Fu, Jingjing Gong and 1 more

CVPR 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 wrist camera in addition to third-person viewOpenVLA-OFT vs OpenVLA-OFT w · LIBERO-Plus, Camera perturbationsuccess rate (%): 10.4 -> 56.4Improved
add confounding objects to layoutpi0, pi0-fast, RIPT-VLA, UniVLA, WorldVLA · LIBERO-Plus, Layoutsuccess rate (%)No clear effect
displace the target object placement and poseVLA models · LIBERO-Plus, Layoutsuccess rate (%)Hurt
mask third-person view only (keep wrist camera) (ablation)OpenVLA-OFT · LIBERO, light ablationsuccess rate (%): 97.1 -> 43.6Hurt
mask both camera views (ablation)OpenVLA-OFT · LIBERO, light ablationsuccess rate (%): 97.1 -> 0.0Hurt
add light perturbation to models with wrist cameraOpenVLA-OFT · LIBERO-Plus, Lightsuccess rate (%): 97.1 -> 88.7No clear effect
perturb language instructionsOpenVLA-OFT · LIBERO-Plus, Languagesuccess rate (%): 97.1 -> 79.5No clear effect
remove the language instruction (blank) (ablation)OpenVLA-OFT · LIBERO object suitesuccess rate (%)No clear effect
remove the language instruction (blank) (ablation)OpenVLA-OFT · LIBERO long suitesuccess rate (%)Hurt
replace the target object in instruction and goal (ablation)OpenVLA-OFT · LIBERO object suite (modified target)success rate (%)Hurt
apply pairwise combined perturbationsOpenVLA-OFT · LIBERO-Plus pairwisejoint success probability: 18.93 -> 18.29 (-0.64)Hurt
fine-tune on generated generalization dataset (20,000+ trajectories)OpenVLA-OFT · LIBERO-Plus, Camerasuccess rate (%): 55.6 -> 92.8Improved

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