VLA Model Training

VLA: Vision-Language-Action

Follow this topic Ask about this topic

Latest papers

55 papers15 papers in the last four weeks, up 150% on the four weeks before. 0.1% of all new papers.

SURE:Spatiotemporal Uncertainty Representation for Efficient Vision-Language-Action Learning
Nov 6, 2026Jingye Zhang, Choo Sin Wai, Dake Zhong +8VLA Model Training Data-Efficient Robot Learning
Tsinghua University
SLA: Scaling Spatial-Language-Action Model via Continual Pre-training
Nov 5, 2026Fuhao Li, Wenxuan Song, Qizhou Huang +7VLA Model Training Spatial Grounding for VLA Models
The Hong Kong University of Science and Technology (Guangzhou) · Beijing Academy of Artificial Intelligence · Tsinghua University +1
Should the Action Expert Train the VLM? Action-Token Supervision and Gradient Routing in Mixture-of-Transformer VLAs
Nov 3, 2026Asher James Hancock, Boyi Li, Yuxiao Chen +6VLA Model Training Flow-Based VLA Models
NVIDIA · Princeton University · Stanford University
AceAction-5M: Hierarchical Goals for Generalist Action Learning Across 3D Game Worlds
Nov 3, 2026Kailing Li, Ruoxuan Zhang, Zheyuan Zhou +5Embodied Agent Benchmarks Vision-Language-Action Models
Tencent · East China Normal University · Jilin University
Vision-Language Models Are Joint Denoisers in Robotic Action Generation
Oct 31, 2026Xicheng Gong, Wenxuan Song, Chengze Liu +5Diffusion-Based VLA Models Vision-Language-Action Models
Peking University · The Hong Kong University of Science and Technology (Guangzhou) · Microsoft Research Asia
BURGER: Benchmarking and Understanding Robotic Generalization via Embodied Drawing
Oct 29, 2026Xiangyu Xu, Bangshuai Peng, Ziyi Liu +5Instruction Generalization in VLA Models Vision-Language-Action Models
Tsinghua University · Shanghai Qizhi Institute
PRIME W0: Physical Priors from World Models for Embodied Manipulation
Oct 29, 2026Huabin Peng, Fagang Jin, Yancong Wei +4Robot Learning Robotic Manipulation
Lumos SZ Lab Team · CUHK · THU +2
PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations
Oct 27, 2026Hanqing Wang, Zhaotian Cai, Ruichuan An +13Vision-Language-Action Models VLA Model Training
Peking University
Vision-Language-Action or World-Action? Just Action Models
Oct 27, 2026Yonghao Yu, Lang Huang, Runyi Li +1VLA Model Training World Action Models
The University of Tokyo · National Institute of Informatics
Evaluation-Guided Demonstration Refinement for Vision-Language-Action Policies: The Winning Entry in the MetaFine Open Competition
Oct 27, 2026Lifan Zheng, Guotao Chen, Zhenyang Li +3VLA Model Training Learning from Demonstration
Southeast University · The Hong Kong Polytechnic University · Monash University +1
RynnVLA-Latent: Scaling Robotic Manipulation Foundation Models with 3D-Grounded Latent Actions
Oct 25, 2026Haoyu Zhao, Xingyue Zhao, Zhian Su +11VLA Model Training Robot Foundation Models
DAMO Academy, Alibaba Group · Hong Kong Embodied AI Lab · CUHK +3
HuGe: Scaling Humanoid Generalist Vision-Language-Action Models with Egocentric Experience
Oct 25, 2026Yang Liu, Jiacheng Liu, Hengtao Li +14VLA Model Training Vision-Language-Action Models
Westlake University · Zhejiang University · Shanghai Innovation Institute +7
TTT in VLA Models: Exploring Training with Human Demonstrations
Oct 23, 2026Qiangong Zhou, Jiafeng Liang, Yimin Dong +2VLA Model Training Vision-Language-Action Models
Standard Robots Co., Ltd.
DRILL: Training World Models to Improve Policies, Not Predict Pixels
Oct 22, 2026Zhaolu Kang, Tailong Luo, Lei Wei +9World Models for VLA Models World Model Learning
Hunyuan, Tencent · Peking University
AQUAACT: A Joystick-Level Underwater Teleoperation Dataset and Closed-Loop Benchmark
Oct 18, 2026Zhenqi Wu, Trung Tien Dong, Aditya Penumarti +5Underwater Robotics VLA Model Training
University of South Florida, Tampa, FL, USA. · University of Florida, Gainesville, FL, USA. · Seoul National University, Seoul, Republic of Korea.
VLA- $Δ$ : Learning Physical State Transition for Vision-Language-Action Models
Oct 15, 2026Qi Lu, Jiarun Li, Ce Zhang +5VLA Model Training Vision-Language-Action Models
Zhejiang University (zju.edu.cn) · Central South University (csu.edu.cn)
Do VLAs Really Follow Instructions? Diagnosing How-Level Instruction Following in Vision-Language-Action Models
Oct 15, 2026Sangwu Park, Wonjoong Kim, Yeonjun In +1Vision-Language-Action Models VLA Model Training
KAIST
What Matters in Quality-Aware VLA Pretraining: A Benchmark and Empirical Study
Oct 13, 2026Mingxuan Yan, Litian Gong, Dihong Huang +19VLA Model Training Vision-Language-Action Models
Georgia Institute of Technology · University of California, Riverside · Axis Robotics +6
Axiom-0: Hierarchical Self-Refinement for Vision-Language-Action Policies
Oct 9, 2026Renbin Shu, Jing Yang, Xiangyuan You +7Vision-Language-Action Models VLA Model Training
4Axiom Robotics Team
Same Scene, Different Task: Skill Alignment for Compositional Generalization in VLAs
Sep 30, 2026Taegeun Yang, Youngju Na, Yoonki Cho +1Vision-Language-Action Models Counterfactual Data Augmentation
Korea Advanced Institute of Science and Technology (KAIST)
DistAL: Distance-based Advantage Learning for VLA Fine-Tuning
Sep 16, 2026Reece O'Mahoney, Ioannis HavoutisReinforcement Learning Reward Shaping
Oxford Robotics Institute, University of Oxford, UK
Decoupling Vision, Language, and Action for Efficient Multi-Task Robot Policies
Sep 16, 2026Xiatao Sun, Chen Liang, Ziyao Zeng +5Visuomotor Policy Learning Efficient VLA Models
Department of Computer Science, Yale University, New Haven, CT, USA. · Peking University, Beijing, China. · Digients, Singapore.
IMLE-VLA: Fast Single-Step Action Generation for Vision-Language-Action Policies
Sep 10, 2026Kian Hosseinkhani, Qinhe Peng, George Shramko +6Efficient VLA Model Inference Vision-Language-Action Models
Simon Fraser University · University of Pennsylvania · Alberta Machine Intelligence Institute +1
HuRo: Robotizing Human Videos for Scalable VLA Pretraining
Sep 9, 2026Jinho Jeong, Se June Joo, Jaehyun Kang +4Cross-Embodiment Robot Learning VLA Model Training
RLWRLD · Yonsei University
ICI-VLA: In-Context Imitation with Spatiotemporally Aligned Demonstrations for Vision-Language-Action Models
Sep 7, 2026Songhua Yang, Ziyu Liu, Xuetao Li +3Imitation Learning Robotic Manipulation
Wuhan University · Institute of Technological Sciences, Wuhan University
Act with Intent: Distilling Behavior Intent for Vision-Language-Action Models
Aug 24, 2026Sangoh Lee, Sangwoo Mo, Wook-Shin HanVLA Model Training Vision-Language-Action Models
GSAI, POSTECH · IME, POSTECH
EXIMO: VLM Guided Exploration of VLA Policies
Aug 20, 2026Bhavya Sukhija, Oliver Groth, Mohit Shridhar +5VLA Model Training Data-Efficient Robot Learning
Google DeepMind
StellaVLA: In-Context Structured Demonstration for Generalizable Vision-Language-Action Models
Aug 12, 2026Siyu Xu, Yunke Wang, Zijian Wang +6Reasoning for VLA Models Continual Learning for VLA Models
StellarEdge AI Technical Team
In-Context VLA: Endowing Vision-Language-Action Models with Language via In-Context Post-Training and Agentic Tool Use
Aug 6, 2026Jiarui Yang, Wen Huang, Jiale Zhang +2Reasoning for VLA Models ICL for Robotics
Nankai University · Tsinghua University · Swiss Federal Institute of Technology in Lausanne (EPFL)
CoTinyVLA: Chain-of-Thought Distillation for a Sub-Billion-Parameter Vision-Language-Action Model
Jul 28, 2026Minhyeok Lee, Chiyoung Kim, Chanhoe Gu +5Reasoning for VLA Models Efficient VLA Model Inference
Chung-Ang University, Republic of Korea

Nov 6, 2026J/K move · Enter open · S save

SURE:Spatiotemporal Uncertainty Representation for Efficient Vision-Language-Action Learning

Jingye Zhang, Choo Sin Wai, Dake Zhong, Zhifang Liu+7

Tsinghua University

VLA Model Training Data-Efficient Robot Learning Uncertainty Estimation for VLA Models Temporal Modeling for VLA Models

Open paper AskPDF arXiv

Vision-language-action (VLA) policies have become a promising paradigm for end-to-end robot control, but their effectiveness still depends heavily on large volumes of demonstration data. In manipulation tasks, this dependence is compounded by structured spatiotemporal uncertainty in the action space: multiple nearby actions or trajectories may successfully solve the same task, while semantically equivalent demonstrations often unfold at different speeds. Standard pointwise supervision treats each demonstrated action as an exact target, imposing an unnecessary learning burden under such uncertainty. We present SURE, a data-efficient VLA training framework that explicitly models both spatial and temporal uncertainty. SURE augments a VLA backbone with a lightweight transposed-convolutional probabilistic action decoder that captures spatial uncertainty through local correlations among neighboring action bins. It further introduces Soft-DTW alignment supervision to accommodate local timing variations between predicted and demonstrated action chunks. Together, these two components reduce sensitivity to demonstrator-specific poses and execution timing, enabling more data-efficient action learning. We evaluate SURE on LIBERO, RoboTwin 2.0, and real-world manipulation tasks. Even with only 10% of the demonstration data or 20% of the training steps, SURE achieves manipulation success rates that match or surpass baseline VLAs.