Vision-Language Models

20VC: The Best AI Companies Have Unique Data Acquisition Strategies | Will Simile Kill Kalshi, Polymarkets and NASDAQ | How to Sign Fortune 500 Companies As Customers in Weeks with Joon Sung Park, Simile

Charts of the Day

AI Micro Dramas, Generative Media, and the Future of Creativity

ContactFlow: A video action conditioning that transfers across embodiments

TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM

πR2: Reactive Real-time Flow Policies

Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?

WOLF-VLA: Whole-Body Humanoid Optimal Locomotion Framework for Vision-Language-Action Learning

DeMaVLA: A Vision-Language-Action Foundation Model for Generalizable Deformable Manipulation

N0-VTLA: Scaling Vision-Tactile-Language-Action Model with Latent Tactile Tokens

Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories

Exp2VLA: Enabling Vision-Language-Action for Drone Navigation from Expert Demonstrations

FELT: Generating Tactile Signals from Vision for Visuo-Tactile Manipulation

LENS: LLM-guided Environment Simplification for Planning and Control in Clutter

Uni-LaViRA: Language-Vision-Robot Actions Translation for Unified Embodied Navigation

147. 和蚂蚁灵波沈宇军聊：机器人原生基础模型、大脑和本体的关系、预训练与数据scale up、老师汤晓鸥

DiMaS: Distribution Matching for Steering Vision-Language-Action Models

DSWAM: A Dual-System World Action Foundation Model for Fine-Grained Robot Manipulation

Patch Policy: Efficient Embodied Control via Dense Visual Representations

Mainframes and Main Characters (This Week in Stratechery)

Reflex: Real-Time VLA Control through Streaming Inference

Demis Hassabis Says AGI Arrives in 2 to 5 Years. Here Is the Full Picture

VIA: Visual Interface Agent for Robot Control

VLAFlow: A Unified Training Framework for Vision-Language-Action Models via Co-training and Future Latent Alignment

DemoBridge: A Simulation-in-the-Loop Toolkit for Single-View Human Demonstration Retargeting

Calf-Integrated Arms for Bimanual Quadruped Loco-Manipulation

Import AI 464: Fables writes GPU kernels; AI automation; and analog computation

Zuckerberg's $14B Bet: Glasses Kill the Phone in 5 Years

S2-VLA: State-Space Guided Vision-Language-Action Models for Long-Horizon Manipulation

ThinkingVLA: Interleaved Vision and Language Reasoning for Robotic Manipulation

Decoupling the Declarative from the Procedural in Vision-Language-Action Models

CoDex: Learning Compositional Dexterous Functional Manipulation without Demonstrations

FurnitureVLA: Learning Long-Horizon Bimanual Furniture Assembly with Vision-Language-Action Model

Building AI for Creators | Luma &amp; Phota Labs

Adapting Generalist Robot Policies with Semantic Reinforcement Learning

Human-as-Humanoid: Enabling Zero-Shot Humanoid Learning from Ego-Exo Human Videos with Human-Aligned Embodiments

VLK: Learning Humanoid Loco-Manipulation from Synthetic Interactions in Reconstructed Scenes

OpenAI Codex lead on the new shape of product work | Andrew Ambrosino

InSight: Self-Guided Skill Acquisition via Steerable VLAs

Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering

PhysVLA: Towards Physically-Grounded VLA for Embodied Robotic Manipulation

OpenHLM: An Empirical Recipe for Whole-Body Humanoid Loco-Manipulation

Co-VLA: Coordination-Aware Structured Action Modeling for Dual-Arm Vision-Language-Action Systems

Pose6DAug: Physically Plausible Multi-view Object Swapping for Robot Data Augmentation

Object-Centric Residual RL for Zero-Shot Sim-to-Real VLA Enhancement

Ideogram’s Open-Weights Image Model and the Future of AI Design

AI, Design, and the Power of Open Models

What Matters in Orchestrating Robot Policies: A Systematic Study of Hierarchical VLA Agents

Hy-Embodied-0.5-VLA: From Vision-Language-Action Models to a Real-World Robot Learning Stack

Improving Robotic Generalist Policies via Flow Reversal Steering

CHORUS: Decentralized Multi-Embodiment Collaboration with One VLA Policy

LARA: Latent Action Representation Alignment for Vision-Language-Action Models

🌈 GLAAD's AI warning

OneVLA: A Unified Framework for Embodied Tasks

ProGAL-VLA: Grounded Alignment through Prospective Reasoning in Vision-Language-Action Models

FlowPRO: Reward-Free Reinforced Fine-Tuning of Flow-Matching VLAs via Proximalized Preference Optimization

GRAIL: Generating Humanoid Loco-Manipulation from 3D Assets and Video Priors

Knowing what your customers want, all the time: Listen Labs' Alfred Wahlforss

ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models

Colosseum V2: Benchmarking Generalization for Vision Language Action Models

VLA-Pro: Cross-Task Procedural Memory Transfer for Vision-Language-Action Models

EXPO-FT: Sample-Efficient Reinforcement Learning Finetuning for Vision-Language-Action Models

FineVLA: Fine-Grained Instruction Alignment for Steerable Vision-Language-Action Policies

Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments

VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies

143. 对何小鹏的第二次访谈：更大赌注、人形机器人Iron诞生、那场意外、技术剧变下CEO、GX和缝合怪

GuidedVLA: Specifying Task-Relevant Factors via Plug-and-Play Action Attention Specialization

Towards Long-horizon Embodied Agents with Tool-Aligned Vision-Language-Action Models

Gemini Robotics — Google IO 2026

Imagine2Real: Towards Zero-shot Humanoid-Object Interaction via Video Generative Priors

Judge, Then Drive: A Critic-Centric Vision Language Action Framework for Autonomous Driving

PokeVLA: Empowering Pocket-Sized Vision-Language-Action Model with Comprehensive World Knowledge Guidance

PointACT: Vision-Language-Action Models with Multi-Scale Point-Action Interaction

MolmoAct2: Action Reasoning Models for Real-world Deployment

StableVLA: Towards Robust Vision-Language-Action Models without Extra Data

165: 英伟达 GEAR 高深远：世界模型、自进化循环、DreamDojo

Do World Action Models Generalize Better than VLAs? A Robustness Study

VLA-ATTC: Adaptive Test-Time Compute for VLA Models with Relative Action Critic Model

Realtime-VLA FLASH: Speculative Inference Framework for Diffusion-based VLAs

TiPToP: A Modular Open-Vocabulary Planning System for Robotic Manipulation

$\Delta$VLA: Prior-Guided Vision-Language-Action Models via World Knowledge Variation

Jim Fan — VLAs Are Dead, WAMs Are Next

LaST-R1: Reinforcing Action via Adaptive Physical Latent Reasoning for VLA Models

Unified 4D World Action Modeling from Video Priors with Asynchronous Denoising

RedVLA: Physical Red Teaming for Vision-Language-Action Models

EmbodiedMidtrain: Bridging the Gap between Vision-Language Models and Vision-Language-Action Models via Mid-training

UniT: Toward a Unified Physical Language for Human-to-Humanoid Policy Learning and World Modeling

Observing and Controlling Features in Vision-Language-Action Models

VP-VLA: Visual Prompting as an Interface for Vision-Language-Action Models

R3D: Revisiting 3D Policy Learning

Vision-Based Safe Human-Robot Collaboration with Uncertainty Guarantees

ViserDex: Visual Sim-to-Real for Robust Dexterous In-hand Reorientation

135. 和自然选择创始人Tristan聊，Elys、赛博分身、灵魂、Context的获取与流动和AI社交网络

HEX: Humanoid-Aligned Experts for Cross-Embodiment Whole-Body Manipulation

LAMP: Lift Image-Editing as General 3D Priors for Open-world Manipulation

Action Images: End-to-End Policy Learning via Multiview Video Generation

SnapFlow: One-Step Action Generation for Flow-Matching VLAs via Progressive Self-Distillation

DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulation

Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models

OpenAI’s Next Image Model Just Leaked. The Examples Are Insane.

Adaptive Action Chunking at Inference-time for Vision-Language-Action Models

MobileManiBench: Simplifying Model Verification for Mobile Manipulation

Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models

ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning

ForeAct: Steering Your VLA with Efficient Visual Foresight Planning

Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution

157: 具身季报26Q1：人形再思考、英伟达世界模型、高自由度灵巧手

Learning Humanoid Navigation from Human Data

DIAL: Decoupling Intent and Action via Latent World Modeling for End-to-End VLA

FocusVLA: Focused Visual Utilization for Vision-Language-Action Models

OmniGuide: Universal Guidance Fields for Enhancing Generalist Robot Policies

SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learning

Rethinking Visual-Language-Action Model Scaling: Alignment, Mixture, and Regularization

ST4VLA: Spatially Guided Training for Vision-Language-Action Models

VLAW: Iterative Co-Improvement of Vision-Language-Action Policy and World Model

Open-World Task and Motion Planning via Vision-Language Model Generated Constraints

Steerable Vision-Language-Action Policies for Embodied Reasoning and Hierarchical Control

Being-0: A Humanoid Robotic Agent with Vision-Language Models and Modular Skills

DualCoT-VLA: Visual-Linguistic Chain of Thought via Parallel Reasoning for Vision-Language-Action Models

WholeBodyVLA: Towards Unified Latent VLA for Whole-Body Loco-Manipulation Control

DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos

Terence Tao – Kepler, Newton, and the true nature of mathematical discovery

Scaling Sim-to-Real Reinforcement Learning for Robot VLAs with Generative 3D Worlds

Sparse Autoencoders Reveal Interpretable and Steerable Features in VLA Models

133. 对谢赛宁的7小时马拉松访谈：世界模型、逃出硅谷、AMI Labs、两次拒绝Ilya、杨立昆、李飞飞和42

Jagdeep Singh (Rhoda AI) — 10 Hours, Not 10,000: $450M to Train Robots on YouTube

Fei-Fei Li (World Labs) — On Creating Large World Models

Fei-Fei Li & Justin Johnson (World Labs) — After LLMs: Spatial Intelligence and World Models

Stanford AI Club: Jeff Dean on Important AI Trends

The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li

How Google’s Nano Banana Achieved Breakthrough Character Consistency

OpenAI Sora 2 Team: How Generative Video Will Unlock Creativity and World Models

Interview with Li Xiang (Part 2): CEO Large Model, MoE, Liang Wenfeng, VLA, Energy, Memory, Confronting Human Nature, Intimate Relationships, Human Wisdom

Interview with Li Xiang Part 2: CEO Large Model, MoE, Liang Wenfeng, VLA, Energy, Memory, Confronting Human Nature, Intimate Relationships, Human Wisdom

138: From the moment you use your mobile phone to when it understands you better, OPPO's mobile phone AI practice | Conversation with Wan Yulong, head of Xiaobu

#38 Karol Hausman & Kevin Black: Building A Brain For Any Robot | AI Eating The Physical World

Fully autonomous robots are much closer than you think – Sergey Levine

【BAAI2025】 Building Physical Intelligence | Karol Hausman

The Race to Create General-Purpose Robots | Karol Hausman & Lachy Groom on TBPN