Robustness of VLA Models

VLA: Vision-Language-Action

This field studies whether robots guided by vision-language-action models can follow instructions and act reliably when viewpoints, lighting, objects, sensor inputs, or robot execution differ from training. Training approaches include perturbation and counterfactual data, consistency objectives, and world-model supervision; inference-time methods use adaptation, monitoring, action correction, and failure recovery. Benchmarks such as LIBERO-Plus and RoboTwin test shifts in simulation, while real-robot evaluations expose hardware faults and contact-related failures; increasingly, evaluations measure safety and behavior over a rollout, not just task completion. A central challenge is improving performance under distribution shift without weakening clean-task performance, sensitivity to meaningful instructions, or safe behavior.

Latest papers

221 papers57 papers in the last four weeks, up 78% on the four weeks before. 0.3% of all new papers.

Tokens of Motion: Primitive Motion Learning for Human-Humanoid Skill Transfer
Oct 27, 2026Hanwen Ren, Meng Zheng, Wenzhe Cui +5VLA Model Adaptation Robustness of VLA Models
United Imaging Intelligence, Boston, MA, USA · Purdue University, West Lafayette, IN, USA
Retrieve to Act: Motion Primitive Graph Retrieval for Robust VLA Execution
Oct 27, 2026Pham Tri Quang, Bao-Ngoc Dao, Minh Duc Nguyen +15Robustness of VLA Models Test-Time Adaptation of VLA Models
Center for AI Research, VinUniversity · VinRobotics · Hanoi University of Science and Technology +6
Late Breaking Results: Audio-Language-Action Models with Direct Speech Conditioning for Robotic Manipulation
Oct 24, 2026Enfu Nan, Pu Zhao, Yixiao Chen +5Language-Conditioned Robot Manipulation VLA Model Adaptation
Northeastern University · EmbodyX Inc.
EgoForesight: Effectively Scaling Policy Learning with Noisy Egocentric Actions
Oct 22, 2026Chubin Zhang, Changyuan Wang, Yirui Wang +7Robot Imitation Learning Robot Learning from Video
Tsinghua University · JD Future Academy · Loopeva +2
FIT-MAC: A Closed-Loop Multi-Agent Framework for Robust Vision-Language-Action Models
Oct 20, 2026Jiatong Li, Jisheng Dang, Qiangqiang Ma +6Closed-Loop Control of VLA Models Robustness of VLA Models
Lanzhou University · Northwest Normal University · Sun Yat-sen University +2
PACT-VLA: Speculative Action-Chunk Verification with Pre-Commit Observations
Oct 20, 2026Yixiang Zhu, Ruilong Wu, Yonghao Chen +3Vision-Language-Action Models Robustness of VLA Models
The Hong Kong University of Science and Technology (Guangzhou) · One Robotics
RoboCasa-Shift: Stage-Aware Robustness Evaluation of Vision-Language-Action Policies in Long-Horizon Household Manipulation
Oct 20, 2026Haocheng Wang, Youquan Liu, Weiyu Guo +6Robustness of VLA Models Robotics
Fudan University · Shenzhen Loop Area Institute, Shenzhen, China · The Chinese University of Hong Kong +2
VLA-Retain: Benchmarking Cross-Dimension Effects of Vision Tuning in VLAs
Oct 20, 2026Borui Xu, Han Chen, Bingsheng He +3VLA Model Adaptation Robustness of VLA Models
National University of Singapore
What Would Removing This Corruption Change? Failure Attribution for Vision-Language-Action Models
Oct 19, 2026Carson Sobolewski, Anjali Parashar, Fei Chen +2Robustness of VLA Models
The authors are with the Massachusetts Institute of Technology, USA.
LIBERO-Pro2.0: A Benchmark for Evaluating the Robustness of Embodied Foundation Models
Oct 17, 2026Xueyang Zhou, Yangming Xu, Qianjiang Li +10Distribution Shift Robustness Embodied Agent Benchmarks
Huazhong University of Science and Technology · Tsinghua University · University of Toronto +2
Enhancing Visual Task-Redirection Attacks on Vision-Language-Action Models via Closed-Loop Optimization
Oct 16, 2026Chunjin Ye, Guanying Chen, Jianjie Huang +6Adversarial Attacks on VLA Models Adversarial Attacks
Sun Yat-sen University, Shenzhen 518107, China · Zhongguancun Academy, Beijing 100094, China · Wuhan University, Wuhan 430072, China +3
CarryOn: A Dual-Loop Harness for Personalized and Reliable Robot Assistance
Oct 15, 2026Zixuan Wang, Yuqi Liu, Yuxin Chen +7Assistive Robotics Robot Safety
HKUST · CUHK · SmartMore
SafeDriveVLA: Navigation-Conditioned World Model Dreaming for Conflict-Aware End-to-End Autonomous Driving
Oct 14, 2026Shaoyuan Xie, Zihan Zhang, Jingxuan Wang +6Autonomous Driving End-to-End Autonomous Driving
UC Irvine · UC San Diego · University of Cincinnati +1
ObjDream: Guiding Vision-Language-Action Models with Object-Level Understanding and Future State Prediction
Oct 14, 2026Huailiang Ma, Yuxuan Li, Ziyi Zhao +7Vision-Language-Action Models Robustness of VLA Models
Southeast University · Shanghai Innovation Institute · Shanghai Jiao Tong University +1
What Matters for World Modeling in Vision-Language-Action Models?
Oct 14, 2026Qisen Ma, Yixiang Chen, Yuan Xu +10World Models for VLA Models Temporal Modeling for VLA Models
New Laboratory of Pattern Recognition (NLPR), Institute of Automation, Chinese Academy of Sciences · School of Artificial Intelligence, University of Chinese Academy of Sciences · Alibaba Group +1
Paraphrase Consensus Is Not Self-Consistency: Instruction Reliance and Robustness in VLA Policies
Oct 14, 2026Hanlong Li, Sihan Tan, Takeshi Ashizawa +2Robustness of VLA Models Instruction Generalization in VLA Models
Institute of Science Tokyo, Tokyo, Japan
End-to-End Vision-Language-Action Policies for Autonomous Lunar Rover Docking
Oct 13, 2026Sitan Huang, Grace GaoSpace Robotics Visuomotor Policy Learning
Stanford University, William F. Durand Building, 496 Lomita Mall, Stanford, CA 94305
LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models
Oct 13, 2026Andrea Yaoyun Cui, Hanyu Wang, Xinyi Wang +6Robustness of VLA Models Robot Manipulation Benchmarks
Luma AI
VLA-GAUGE: Systematically Evaluating Generalization and Robustness in Vision-Language-Action Models
Oct 12, 2026Jun Xiao, Rui Ye, Dong Cao +3Robustness of VLA Models
Software College, Northeastern University, Shenyang, China · School of Computer Science, Shanghai Jiao Tong University, Shanghai, China · Shanghai Syslong Information Technology Co., Ltd.
Same Goal, Different Words: Spatial Instruction Sensitivity in World-Action Models
Oct 12, 2026Ali-Adeeb Abbas, Esmaeil Seraj, Behrad Toghi +1World Model Evaluation Vision-Language-Action Models
RIVeR Laboratory, Northeastern University, Boston, MA, USA. · General Motors Robotics, Mountain View, CA, USA.
Barrier-Constrained Guidance For Real-time Asynchronous VLA Execution
Oct 10, 2026Wenbo Yu, Jifan Li, Wenke Xia +5Inference-Time Guidance Efficient VLA Model Inference
Beijing Forestry University · Beijing Innovation Center of Humanoid Robotics · Gaoling School of Artificial Intelligence, Renmin University of China
Red-Teaming of Vision-Language-Action Models via Quality Diversity Environment Generation
Oct 10, 2026Shihan Zhao, Henry Chen, Ya-Chuan Hsu +7Robustness of VLA Models Adversarial Attacks on VLA Models
Thomas Lord Department of Computer Science, University of Southern California. · Columbia University. · Adobe Research +1
JitterVLA: A Replay Benchmark for Motion Quality in VLA-Driven Human-Robot Manipulation
Oct 9, 2026Faith Ogunwoye, Didar Seyidov, Konstantin Gubernatorov +1Robotic Control Robotic Policy Evaluation
Intelligent Space Robotics Laboratory, Skolkovo Institute of Science and Technology Moscow, Russia
RARA: Robust Policy Fine-Tuning through Retrieval-Anchored Representation Alignment for Robotic Manipulation
Oct 9, 2026Xuhui Kang, Sung-Wook Lee, Yan Zhuang +2Robot Policy Adaptation Robustness of VLA Models
University of Virginia
OGPO: Offline Goal-conditioned Policy Optimization for Recoverable Vision-Language-Action Models
Oct 9, 2026Xule Gao, Xi Wang, Zehua Zang +3Vision-Language-Action Models Robotic Manipulation
National Key Laboratory of Space Integrated Information System · Institute of Software, Chinese Academy of Sciences · University of Chinese Academy of Sciences +1
VersaCamVLA: Camera-Configurable VLA Policies for Robotic Manipulation
Oct 8, 2026Boyao Han, Chen Shi, Jingjing Qian +2Vision-Language-Action Models Robotic Manipulation
The Chinese University of Hong Kong, Shenzhen · Shenzhen Loop Area Institute · Harbin Institute of Technology, Shenzhen
CAPABLE: Capability-Aware Policy Adaptation via Behavioral Latent Encoding
Oct 8, 2026Mohammad Khoshnazar, Mohammad Dehghani Tezerjani, Deyuan Qu +6Robot Policy Adaptation Robot Failure Recovery
University of Bremen, Bremen, Germany. · University of North Texas, Denton, TX, USA. · Toyota Motor North America, InfoTech Labs, Mountain View, CA, USA.
WARP-VLA: Wrist-Camera Adaptation for View-Robust Policy Execution in Vision-Language-Action Models
Oct 8, 2026Junmyeong Lee, Dongmin Shin, Min-Gyu Park +3Robustness of VLA Models Vision-Language-Action Models
Yonsei University. · Korea Electronics Technology Institute(KETI).
When Listening Becomes Easier: Scrubbing Visual Cues for Shortcut-Free VLAs
Oct 7, 2026Jasper Gerigk, Kenzo Aspuru-Takata, Chin-Hsuan Wu +3Vision-Language-Action Models Robustness of VLA Models
University of Toronto, Toronto, ON M5S 3H5, Canada · Vector Institute, Toronto, ON M5G 0C6, Canada
Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models
Oct 7, 2026Mikey Watts, Yuchen CuiRobustness of VLA Models Instruction Generalization in VLA Models
Independent Researcher · University of California, Los Angeles

Oct 27, 2026J/K move · Enter open · S save

Tokens of Motion: Primitive Motion Learning for Human-Humanoid Skill Transfer

Hanwen Ren, Meng Zheng, Wenzhe Cui, Benjamin Planche+4

United Imaging Intelligence, Boston, MA, USA · Purdue University, West Lafayette, IN, USA

VLA Model Adaptation Robustness of VLA Models Robotic Manipulation Flow-Based VLA Models

Open paper AskPDF arXiv

Vision-language-action (VLA) models provide a promising paradigm for scalable robotic manipulation, yet their reliance on success-only behavioral cloning leaves them brittle; lacking corrective training signals, minor execution errors rapidly compound into unrecoverable, out-of-distribution failures. To address this limitation, we propose Adaptive Failure-Informed Learning (AFIL), an end-to-end framework that leverages failure trajectories as adaptive negative guidance for diffusion- and flow-based VLA policies. AFIL uses a pretrained VLA to generate failure rollouts online, avoiding the need for handcrafted failure-mode design or human-in-the-loop recovery. It then jointly trains Dual Action Generators (DAGs) for successful and failed behaviors while sharing a common vision-language backbone, enabling efficient failure-aware policy learning with limited parameter overhead. During sampling, the failure generator adaptively steers action generation away from failure-prone regions and toward more reliable success modes, with guidance strength determined by the per-diffusion-step distance between success and failure distributions. Experiments across in-domain and out-of-domain robotic manipulation tasks, covering both short- and long-horizon settings, show that AFIL consistently improves task success rates and robustness over existing VLA baselines, demonstrating its effectiveness, efficiency, and generality.