Efficient VLA Model Inference

VLA: Vision-Language-Action

This topic studies how to run vision-language-action models fast enough for responsive, reliable robot control, including on edge devices. Common approaches prune or cache visual tokens, quantize or distill models, reduce action-generation steps, and overlap inference with observation and execution. Methods are tested on manipulation benchmarks such as LIBERO and RoboTwin, as well as real robots, using measures like latency, control frequency, memory, energy, and task success. The main tension is that faster inference can undermine closed-loop reliability or motion quality, so gains must be assessed in full task execution rather than by model speed alone.

Latest papers

190 papers56 papers in the last four weeks, up 115% on the four weeks before. 0.3% of all new papers.

Recursive Belief Vision Language Action Models
Feb 24, 2026Vaidehi Bagaria, Bijo Sebastian, Nirav Kumar PatelVision-Language-Action Models Long-Horizon Robotic Manipulation
Department of Engineering Design, Indian Institute of Technology, Madras, Chennai, India
BFA++: Hierarchical Best-Feature-Aware Token Prune for Multi-View Vision Language Action Model
Feb 24, 2026Haosheng Li, Weixin Mao, Zihan Lan +6Efficient VLA Model Inference Visual Token Pruning
Institute of Software, Chinese Academy of Sciences, Beijing, China. · University of Chinese Academy of Sciences, Beijing, China. · LimX Dynamic, Shenzhen, China. +3
QuantVLA: Scale-Calibrated Post-Training Quantization for Vision-Language-Action Models
Feb 23, 2026Jingxuan Zhang, Yunta Hsieh, Zhongwei Wan +5Efficient VLA Model Inference VLA Model Post-Training
The Ohio State University · University of Michigan · City University of Hong Kong
Global Prior Meets Local Consistency: Dual-Memory Augmented Vision-Language-Action Model for Efficient Robotic Manipulation
Feb 22, 2026Zaijing Li, Bing Hu, Rui Shao +5Efficient VLA Model Inference Memory-Augmented VLMs
Harbin Institute of Technology, Shenzhen · PengCheng Laboratory · Shenzhen Loop Area Institute +1
Habilis- $β$ : A Fast-Motion and Long-Lasting On-Device Vision-Language-Action Model
Feb 21, 2026Tommoro Robotics, :, Jesoon Kang +22Vision-Language-Action Models Robotics
Tommoro Robotics
How Fast Can I Run My VLA? Demystifying VLA Inference Performance with VLA-Perf
Feb 20, 2026Wenqi Jiang, Jason Clemons, Karu Sankaralingam +1Efficient VLA Model Inference On-Device Inference
NVIDIA Research
DynamicVLA: A Vision-Language-Action Model for Dynamic Object Manipulation
Jan 29, 2026Haozhe Xie, Beichen Wen, Jiarui Zheng +4Inference-Time Optimization Efficient VLM Inference
Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
Oct 9, 2025Jason Jabbour, Dong-Ki Kim, Max Smith +6Vision-Language-Action Models Efficient VLA Model Inference
FieldAI · Harvard University
AdaThinkDrive: Adaptive Thinking via Reinforcement Learning for Autonomous Driving
Sep 17, 2025Yuechen Luo, Fang Li, Shaoqing Xu +11Autonomous Driving Reasoning for VLA Models
Tsinghua University · Xiaomi EV · Nanyang Technological University +2
VOTE: Vision-Language-Action Optimization with Trajectory Ensemble Voting
Jul 7, 2025Juyi Lin, Amir Taherin, Arash Akbari +11Robotic Control Efficient VLA Model Inference
Northeastern University, Boston, USA · EmbodyX,San Mateo,USA

Feb 24, 2026J/K move · Enter open · S save

Recursive Belief Vision Language Action Models

Vaidehi Bagaria, Bijo Sebastian, Nirav Kumar Patel

Department of Engineering Design, Indian Institute of Technology, Madras, Chennai, India

Vision-Language-Action Models Long-Horizon Robotic Manipulation VLMs for Robotics Closed-Loop Control of VLA Models Temporal Modeling for VLA Models Efficient VLA Model Inference+2

Open paper AskPDF arXiv

Vision-language-action models must enable agents to execute long-horizon tasks under partial observability. However, most existing approaches remain observation-driven, relying on short context windows or repeated queries to vision-language models (VLMs). This leads to loss of task progress, action repetition under perceptual aliasing, and high inference latency. While semantic grounding is important, long-horizon manipulation fundamentally requires persistent, action-conditioned state representations. Current VLAs lack such representations and exhibit limited temporal and physical reasoning, making them ill-suited for multi-stage control. This paper introduces RB-VLA, a belief-centric architecture trained with self-supervised world-model objectives that maintains a compact latent state encoding task-relevant history, dynamics, and object interactions. Queried once per task, the VLM provides high-level intent, while the belief tracks task progress and enables phase-aware, causally grounded control under partial observability without storing raw observations or scaling memory with time. The belief and intent jointly condition a diffusion policy for robust closed-loop execution. RB-VLA outperforms prior VLAs on long-horizon benchmarks, achieving 52.5 percent and 37.5 percent higher success rates on multi-stage pick-and-place and stacking tasks, respectively, compared to pi_0. It also reduces inference latency by up to five times relative to baselines and eliminates memory growth across timesteps observed in existing VLAs. Ablations show the belief module is the primary driver of performance, increasing success rates from 32.5 percent without belief to 77.5 percent with belief.