VLA Model Adaptation

VLA: Vision-Language-Action

Follow this topic Ask about this topic

Latest papers

48 papers22 papers in the last four weeks, up 175% on the four weeks before. 0.1% of all new papers.

PRISM: Primitive Routing via In-context Skill Mixing for Retentive VLA Adaptation
Nov 6, 2026Jiachen Tao, Gengyu Zhang, Haoxuan Wang +2VLA Model Adaptation VLMs for Robotics
University of Illinois Chicago
Human-to-Robot Transfer via Bridging Representation and Action Space in VLAs
Nov 3, 2026Junhyeong Park, Wook Jung, Taeyoung Kim +2VLA Model Adaptation Cross-Embodiment Robot Learning
KAIST · RLWRLD
RAFT: Reasoning-Aligned Fine-Tuning Improves Robustness of VLA Post-Training
Nov 3, 2026Shresth Grover, Andrey Kolobov, Dhruv ShahVLA Model Adaptation Vision-Language-Action Models
1Princeton University · 2Microsoft Research
Controlling Action-Head Co-Adaptation to VLM Representation Anisotropy
Nov 2, 2026Ruijie Zhang, Jiayin Zheng, Yequan Zhao +8VLA Model Adaptation Representation Geometry
University of California, Santa Barbara · Foundation Model Team, XPeng
FineCraft: Enhancing VLA Models for Fine-Grained Manipulation
Nov 2, 2026Shuaike Zhang, Ziheng Quan, Junhao Ran +3Robotic Manipulation Language-Conditioned Robot Manipulation
Harbin Institute of Technology, Shenzhen · Xi’an Jiaotong University · Shenzhen Loop Area Institute HITSZ Robotics
Behavioral Steering of Vision-Language-Action Models in the Low-Parameter Regime
Oct 31, 2026James Tcheng, Feiyu Gavin ZhuVLA Model Adaptation RL for VLMs
Carnegie Mellon University
CalVLA: Shaping Native Confidence in Vision-Language-Action Policies with Outcome Supervision
Oct 30, 2026Boniface Chukwuemeka GodwinFailure Detection for VLA Models Confidence Estimation for VLMs
Carnegie Mellon University Africa, Kigali, Rwanda
Improving Fine-Grained Manipulation through Demonstration Expansion and Execution Design
Oct 29, 2026Jingxu Gu, Dongze LianRobotic Manipulation VLA Model Adaptation
Nanjing University
SaiVLA-JEPA: Evaluating Future Latent Prediction in JEPA-Conditioned VLA Control
Oct 28, 2026Rufeng Xiao, Ruihao Xiao, Wei Chen +1Closed-Loop Control of VLA Models Temporal Modeling for VLA Models
VLA Research Team, Synthoid.ai
Self-Labeled Preference Transfer for Steerable Robot Manipulation
Oct 27, 2026Kaiwen Hong, Shaoting Peng, Xiangyu Zeng +6VLA Model Adaptation Robotic Manipulation
University of Illinois Urbana-Champaign · University of California, Berkeley · Harvard University +2
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
JevVLA: Jev-Inspired Energy-Based Adaptation for Vision-Language-Action Policies
Oct 26, 2026Wei Wang, Wenqiao Zhang, Zheqi Lv +6VLA Model Adaptation Robotic Manipulation
College of Computer Science and Technology, Zhejiang University · Cornell University · Universal Ubiquitous AI Co., Ltd. +2
X-World: Transferring Video Priors to Variable-View Robot Control
Oct 25, 2026Chun Gu, Xiaofei Wei, Yiming Chen +2Robotic Control Vision-Based Robot Control
School of Data Science, Fudan University · Shanghai Innovation Institute · Logos Intelligence Lab
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.
When Human Corrections Travel: Safety in Federated VLA Adaptation
Oct 23, 2026Vigneshwaran Subbaraju, Obaidullah Zaland, Monowar BhuyanVLA Model Adaptation Human-in-the-Loop AI
A*STAR, Singapore · Ume˚a University, Sweden
MAST-VLA: Skill-Conditioned Adaptation with Trajectory-Affordance Priors for Multi-Skill Robot Manipulation
Oct 23, 2026Jiachi Yang, Zhiyi Xu, Hongquan Zhang +4Multi-Task Robotic Manipulation VLA Model Adaptation
Shanghai Innovation Institute · Tianjin University · Tongji University +1
From Tabletop to Walking Bot: Deploying a VLA Model on a Legged Manipulator Robot
Oct 23, 2026Giovanni Minelli, Antonio Langella, Giovanni B. Dessy +4Loco-Manipulation VLA Model Adaptation
Dynamic Legged Systems Lab, Istituto Italiano di Tecnologia (IIT)
Full-Weight VLA Fine-Tuning Without an fp32 Master Copy Silently Erases Most Updates
Oct 22, 2026Shubham SinghVLA Model Adaptation Efficient VLA Models
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
HAF-VLA: Hierarchical Action Flow Framework for Humanoid Vision-Language-Action Control
Oct 19, 2026Langzhe Gu, Chengkai Hou, Meng Li +12Humanoid Robot Control Humanoid Loco-Manipulation
State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University · Beijing Innovation Center of Humanoid Robotics · Nankai University +2
Retrieval-triggered Action Projection Adjustment for Frozen VLA Policies: An Empirical Study and Failure Analysis
Oct 17, 2026Jinming Wang, Ruiheng Zhang, Yong Wang +2Vision-Language-Action Models VLA Model Adaptation
Department of Computer Science and Technology, Hainan Bielefeld University of Applied Sciences, Danzhou, China
Look Before You Act: Bridging Active Perception and Vision-Language-Action Models via Observation-Aligned Data Adaptation
Oct 16, 2026Jiawei Geng, Zhenyang Liu, Yuzheng Wu +2Active Vision for Robot Manipulation VLA Model Adaptation
Fudan University · Shanghai Innovation Institute
One-shot Human Video-Guided Adaptation for Vision-Language-Action Models
Oct 14, 2026Qi Chen, Yuwei Wu, Quan Chen +5VLA Model Adaptation Vision-Language-Action Models
Beijing Key Laboratory of Intelligent Information Technology, School of Computer Science and Technology, Beijing Institute of Technology. · Guangdong Laboratory of Machine Perception and Intelligent Computing, Shenzhen MSU-BIT University. · AInnovation Technology Group Co., Ltd.
Post-Training, Finetuning, and Steering:A Field Guide for Curating Training Data
Oct 13, 2026Maria Attarian, Yifan Ruan, Claas Voelcker +5VLA Model Adaptation RL Fine-Tuning
University of Toronto · University of Texas at Austin
Verify, Act, Distill: Turning Test-Time Compute into Test-Time Adaptation for Vision-Language-Action Models
Oct 10, 2026Kaneyoshi Hiratsuka, Satoshi Yamamori, Jun MorimotoTest-Time Adaptation of VLA Models VLA Model Adaptation
Kyoto University
Noise Space Learning for VLA Adaptation
Oct 9, 2026Junjie Lu, Chuheng Zhang, Yucen Wang +4VLA Model Adaptation Vision-Language-Action Models
University of Technology Sydney · Microsoft Research · Nanjing University +1
VLA-ZO: Fast Zeroth-Order Adaptation for Vision-Language-Action Models
Oct 5, 2026Jaemin Kim, Jiahn Kim, Taesik GongZeroth-Order Optimization Efficient VLA Model Inference
UNIST
When Does Retrieval Help? A Study of In-Context Adaptation in Vision-Language-Action Models
Oct 4, 2026Zixuan Liu, Joris Köster, Zizhan Zheng +1Vision-Language-Action Models Retrieval-Augmented ICL
Department of Computer Science, Tulane University · Department of Computer Science, Aalto University · Department of Industrial Engineering and Management, Aalto University
Inline Memory Meets Reusable Skills: Memory-centric Framework for Vision-Language-Action Model
Sep 30, 2026Zaijing Li, Rui Shao, Bing Hu +3Memory-Augmented VLMs VLA Model Adaptation
Harbin Institute of Technology (Shenzhen) · Pengcheng Laboratory
Data-Efficient Adaptation of a Driving VLA to Class 8 Trucks
Sep 29, 2026Satyajeet Das, Aaron Buxbaum, Niels Joubert +1VLA Model Adaptation Autonomous Driving
Department of Computer Science, University of Southern California, Los Angeles, CA 90089, USA · Stack AV, Pittsburgh, PA, USA

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

PRISM: Primitive Routing via In-context Skill Mixing for Retentive VLA Adaptation

Jiachen Tao, Gengyu Zhang, Haoxuan Wang, Bing Liu+1

University of Illinois Chicago

VLA Model Adaptation VLMs for Robotics Continual Learning for VLA Models

Open paper AskPDF arXiv

Vision-Language-Action (VLA) models promise a single language-conditioned policy adaptable to many manipulation behaviors, yet deployment additionally requires skill retention and long-horizon composition: a policy must retain previously acquired skills and reuse them inside composite tasks whose behavior changes across phases. The standard recipe absorbs every new task into one shared update, which provides neither a parameter-level locus for reusable skills nor a mechanism for invoking each skill along the trajectory. We observe that this missing structure is latent rather than absent: per-skill low-rank updates of a pretrained VLA occupy a robustly distinguishable, low-dimensional subspace, and a single composite-task LoRA concentrates roughly half of its Frobenius energy in this subspace, against a near-zero matched random-rotation control. Building on this geometry, we introduce PRISM, which represents reusable skills as parameter-level primitives: each skill family is stored as a dedicated low-rank action-expert adapter in a primitive bank, primitives are augmented with skill-specific low-rank residuals and composed at inference by a per-chunk VLM-conditioned router that reads the prefix representation to track task progress, and the router is supervised by per-chunk labels from a deterministic, task-conditioned, action-derived oracle. During composite training, the vision encoder, the language LoRA, the skill residuals, the router, and the shared action projections and time MLP are updated; every primitive is held frozen, so the base per-skill action-expert adapters are preserved by construction. On RoboCasa365, PRISM achieves 56.43% atomic-skill success and reaches 20.00% on long-horizon composite tasks, while a standard

π_{0.5}

-LoRA two-stage fine-tune reaches 17.86% atomic and 17.50% composite success. We additionally validate PRISM on an AgileX PIPER real-robot setup.