All Episodes
Embodied AI 101 — 387 episodes · page 1 of 4
Motion- and Gravity-Aware 3D Reconstruction: Do Offline Priors Help?
ASPIRE: Can Code Refinement Make Robot Manipulation Generalize?
UAV-ESPrune: Can Visual Token Pruning Preserve UAV Reasoning?
Gemini Robotics 2: How General Is Its Whole-Body Robot Control?
RoboJEPA to Long-WAM: Do New Embodied AI Models Generalize?
VLA-ACL: Can Token Pruning Speed Up Robot Manipulation?
Juno: Do Predictive Latents Make VLA Manipulation More Robust?
Can Rephrasing Fix VLA Language Sensitivity Without Retraining?
Can CriticGPT Replace Hand-Designed Rewards for Robot Manipulation?
Urbex: Can Agentic VLM Search Improve City-Scale 3D Grounding?
VPP2: Does Better Video Prediction Improve Robot Manipulation?
PointWAM: Does 3D World Modeling Improve Dexterous Manipulation?
Mulligan: Can Failure-Focused Data Improve Robot Imitation Learning?
SOLE-R1: Can Video-Language Rewards Train Manipulation Policies?
Odyssey-3: Does a Physics-IQ World Model Transfer to Real Robots?
Long-WAM: Does Longer Visual Context Improve Real-Time Robot Control?
WR1.0: Does Whole-Body Humanoid Control Prove Generalization?
WorldGuide: Better Video Planning, Not Proven Robot Control
R2SGen: Can Constraint Solving Improve Real-to-Sim Driving Scenarios?
Humanoid Scooter Control: Does Sim-to-Real Reinforcement Learning Work?
ENPIRE: Can Coding Agents Improve Real-World Robot Manipulation?
GAS-Robo: Can Grid Actions Improve LLM-Driven Manipulation?
Four Keypoints, One Careful Pour: Connecting Language to Precise Robot Motion
PEAfowl: Helping Two Arms Agree on What They See
MotorMind: Giving a Frozen VLM the Controls
MAPF-World: Why a Little Foresight Can Beat a Bigger Policy
Digit Leaves the Lab: The Real Lesson of 500 Demonstrations
VLS: Steering Frozen Robot Policies Without Retraining
CAST: Making Language Matter to Robot Policies
When the Command Disappears: What Counts as Robot Intelligence?
Fast Plans Are Not Enough: Does the Robot Actually Use Its Plan?
Atlas's New Hands: Built to Learn, Not Just to Look Human
Astra Gets a Body: What Zero-Shot Humanoid Control Actually Transfers
Physical Coding: Robots That Check Their Work—and Keep the Fixes
SpatialClaw: Give the VLM a Geometry Notebook
Real-Time EXPO-FT: Slow VLA, Fast Corrections, Better Control
Aether: From Plausible Futures to Feasible Robot Plans
ARLI: Fixing the Missing State in Asynchronous Robot RL
Flex-π: Teach Geometry, Choose How Much to Imagine
SkipVLA: Let the Planner Move, Let the VLA Manipulate
ForeDrive: A World Model Is Only Useful If It Changes the Plan
DO AS I DO: Making Human Video Executable
GAM: Ground the Target Before You Learn the Action
Tune Slowly, Control Quickly: Learning a Better Robot Navigation Stack
THAW-VLA: World-Model Features Without World-Model Latency
Joga: Why a Soccer Humanoid Needs to Control Its Gaze
AthenaZero: Why Dynamic Manipulation Starts With Lower Inertia
FROA-Drive and the Hard Part of Targeted Policy Repair
Teach the Robot Where to Act—Then Make It Fast
HEAR: The Sound Your Robot Missed Between Actions
VS-Splat: Learning Where 3D Detail Belongs
Spatial Search Comes to Quest — Beyond Pretty Gaussian Splats
Occamy-1.0: Teaching a 35B Agent to Finish the Job
Helix 2.5: The House Changed. The Policy Didn't.
The Hard Part Was the Stack: PI's Robot Goes to Work
ZDTaichu5.0-9B: Better Spatial Reasoning, an Unfinished Edge Story
Light-Loco-Parkour: Learning the Skill Is Only Half the Problem
OM-1: Human Skills, Many Bodies
Show-Harness Gives VLMs a Robot Keyboard
BridgeVLA++: Keep the Geometry, Give the Robot a Memory
The Sim-to-Real Gap Inside the Gearbox
A Stable Grasp Isn't Enough: Giving 6-DoF Grasping a Purpose
HiDream-O1-Embodied: A World Model Must Earn Its Actions
Axis's Real2Sim2Real Bet: Make the Data Prove Itself
A Rubik's Cube, Two Robot Hands, and One Category Error
UnifoLM: Predict the Interaction, Not the Whole World
Stop Making the Video Model Remember the World
Atlas Turns Phone Scans Into Robot Worlds—But Not Yet Into Physics
W²G-Net: Carrying Instance Identity from Shiny Scrap to the Gripper
When Avoidance Is Not Enough: T-CARE Gives UAV Swarms a Right-of-Way
Teaching Robots What "Better" Means
TANGO: Stop Treating a Humanoid Like a Disk on the Floor
Stop Teaching the VLA Your Camera: Robot-Centric Pointmaps as an Action-Aligned Visual Interface
One Policy, Many Bodies: Qwen-VLA's Bid to Unify Embodied AI
When the Simulator Lies About Friction: Physics as a Transfer Prior for Touch
Seeing Around Corners, Acting at the Edge: REACT as a VLA Safety Co-Pilot
RoboTok: Turning the Internet Into a Motion Search Engine for Dexterous Robots
VLAct: The VLA Backbone Is Not Just Plumbing
Two Models, One Gradient: Making World-Model RL Work for Contact-Rich Robots
GigaBrain-0.7: Scaling Robot Intelligence by Separating Acting, Thinking, and Imagining
Geometry Is the Contact Plan: Bringing Surface Reasoning Into MPC
A Swarm That Knows Its Margin: Safe Multi-Drone Transport With CBF-RL
The Optimizer Becomes a Policy: Inside RP-1's Learned Planner for Robot World Models
SAGA: Ground the Affordance, Then Learn the Motion
REMAC: Turning Robot Failures into Better Team Plans
Restore the Sensor Contract: Training-Free Robustness for Brittle VLA Policies
Before the Robot Touches Anything: Physics-Grounded Deliberation with Embodied Tree of Thoughts
MemoAct: Why Robot Memory Needs Both a Scratchpad and an Archive
When Touch Means Something: OmniVTLA and the Missing Contact Layer in Robot Foundation Models
Gaussians as State, Futures as Supervision: Inside GSP-3D
SIRModel and the Missing Geometry Layer in Vision–Language Manipulation
The Robot That Knows When to Think Twice: Inside τ0-VLA
When Should a Robot Stop Imagining? RISE and Adaptive World-Action Planning
The Crib as a Causal Laboratory: How a Playing iCub Learns "I Did That"
Feeling Around Corners: STRP Compresses Fingertip Force into Directional Touch
Give the Swarm a Pheromone Compass
When Every Weld Looks the Same: Language-Labeled Graphs for Industrial Localization
When Robot Eyes Converge: Active Stereo for Humanoid Depth
Two Clocks for Touch: How T-Rex Gives Dexterous VLAs a Reflex Arc
Let the VLM Decide, Let the Planner Drive: Inside Senna