VLA
DropVLA: An Action-Level Backdoor Attack on Vision--Language--Action Models
VLA 安全攻击评估，提供动作级后门测试工具
cs.RO
閱讀原文
VLA 研究日報 Pulsar
LIVE
— AI 線今日無資料 —— AI 线今日无资料 —
VLA 線VLA 线 · cs.RO · cs.AI · cs.LG
- Beyond State-Wise Mirror Descent: Offline Policy Optimization with Parameteric Policies 离线 RL 理论分析，无明确 VLA 架构或机器人实验 CS.AI
- DropVLA: An Action-Level Backdoor Attack on Vision--Language--Action Models VLA 安全攻击评估，提供动作级后门测试工具 CS.RO
- DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulation VLA 推理加速，动态静态层跳过，可本周集成 CS.RO
- Point Bridge: 3D Representations for Cross Domain Policy Learning [NVIDIA|Pinto] 3D 表征跨域策略，Pinto/NVIDIA 出品，高潜 CS.RO
- RoboMIND 2.0: A Multimodal, Bimanual Mobile Manipulation Dataset for Generalizable Embodied Intelligence 双臂移动操作数据集，重要但不紧急 CS.RO
- SocialNav: Training Human-Inspired Foundation Model for Socially-Aware Embodied Navigation 社交导航策略，方法相邻但无明确 VLA 路径 CS.RO
- Attentive Feature Aggregation or: How Policies Learn to Stop Worrying about Robustness and Attend to Task-Relevant Visual Cues 视觉特征聚合模块，方法创新一般 CS.RO
- Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward RL 框架改进，非 VLA 架构直接相关 CS.RO
- AoE: Always-on Egocentric Human Video Collection for Embodied AI 第一人称视频数据集，重要但不紧急 CS.RO
- Robust Skills, Brittle Grounding: Diagnosing Restricted Generalization in Vision-Language Action Policies via Multi-Object Picking VLA 泛化性诊断分析，无新策略提出 CS.RO
- Planning from Observation and Interaction 规划方法，标题通用，无 VLA 具体贡献 CS.RO
- ABPolicy: Asynchronous B-Spline Flow Policy for Real-Time and Smooth Robotic Manipulation 策略平滑化方法，非 VLA 核心架构 CS.RO
VLA 研究日報VLA 研究日报
共 20 篇
🔧 技術技术
VLA
DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulation
VLA 推理加速，动态静态层跳过，可本周集成
cs.RO
閱讀原文
VLA [NVIDIA|Pinto]
Point Bridge: 3D Representations for Cross Domain Policy Learning
3D 表征跨域策略，Pinto/NVIDIA 出品，高潜
cs.RO
閱讀原文
VLA
FAVLA: A Force-Adaptive Fast-Slow VLA model for Contact-Rich Robotic Manipulation
触觉 VLA 新架构，力自适应快慢系统，接触操作
cs.RO
閱讀原文
VLA
QuantVLA: Scale-Calibrated Post-Training Quantization for Vision-Language-Action Models
VLA 量化部署，尺度校准后训练量化，易用
hf-papers
閱讀原文
📖 背景閱讀背景阅读
VLA
Beyond State-Wise Mirror Descent: Offline Policy Optimization with Parameteric Policies
离线 RL 理论分析，无明确 VLA 架构或机器人实验
cs.AI
閱讀原文
VLA
RoboMIND 2.0: A Multimodal, Bimanual Mobile Manipulation Dataset for Generalizable Embodied Intelligence
双臂移动操作数据集，重要但不紧急
cs.RO
閱讀原文
VLA
SocialNav: Training Human-Inspired Foundation Model for Socially-Aware Embodied Navigation
社交导航策略，方法相邻但无明确 VLA 路径
cs.RO
閱讀原文
VLA
Attentive Feature Aggregation or: How Policies Learn to Stop Worrying about Robustness and Attend to Task-Relevant Visual Cues
视觉特征聚合模块，方法创新一般
cs.RO
閱讀原文
VLA
Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward
RL 框架改进，非 VLA 架构直接相关
cs.RO
閱讀原文
VLA
AoE: Always-on Egocentric Human Video Collection for Embodied AI
第一人称视频数据集，重要但不紧急
cs.RO
閱讀原文
VLA
Robust Skills, Brittle Grounding: Diagnosing Restricted Generalization in Vision-Language Action Policies via Multi-Object Picking
VLA 泛化性诊断分析，无新策略提出
cs.RO
閱讀原文
VLA
Planning from Observation and Interaction
规划方法，标题通用，无 VLA 具体贡献
cs.RO
閱讀原文
VLA
ABPolicy: Asynchronous B-Spline Flow Policy for Real-Time and Smooth Robotic Manipulation
策略平滑化方法，非 VLA 核心架构
cs.RO
閱讀原文
VLA
Hybrid Offline-Online Reinforcement Learning for Sensorless, High-Precision Force Regulation in Surgical Robotic Grasping
手术机器人力控，应用场景过窄
cs.RO
閱讀原文
VLA
OmniXtreme: Breaking the Generality Barrier in High-Dynamic Humanoid Control
标题夸大通用性，实验可能仅限仿真 [💧灌水]
cs.RO
閱讀原文
VLA
StemVLA:An Open-Source Vision-Language-Action Model with Future 3D Spatial Geometry Knowledge and 4D Historical Representation
标题夸大 3D/4D 概念，无具体技术路线 [💧灌水]
cs.RO
閱讀原文
VLA
VCA: Vision-Click-Action Framework for Precise Manipulation of Segmented Objects in Target Ambiguous Environments
点击交互 VLA 变体，应用场景较窄
cs.RO
閱讀原文
VLA
Demystifying Action Space Design for Robotic Manipulation Policies
动作空间设计分析，无新策略提出
cs.RO
閱讀原文
VLA
SignVLA: A Gloss-Free Vision-Language-Action Framework for Real-Time Sign Language-Guided Robotic Manipulation
手语引导 VLA，应用特定非核心架构
hf-papers
閱讀原文