Shows / Embodied AI 101
Embodied AI 101
Embodied AI 101 tracks the latest research at the intersection of AI and physical intelligence. Robotics, manipulation, world models, and the path from digital intelligence to embodied agents — each episode breaks down recent papers and advances in the field. We examine the methodology, real-world implications, and path to deployment. Covering robot learning, perception, control, and the fundamental challenge of grounding intelligence in the physical world. New episodes daily.
Episodes
- Embodied AI 101
LT2: Linear-Time Looped Transformers
Replaces quadratic softmax attention in looped architectures with linear/sparse mechanisms for iterative memory refinement, achieving parity with standard…
- Embodied AI 101
One Learning Rate Doesn't Fit All: Layerwise Spectral Scheduling for Transformers
Shows that modern transformers are highly heterogeneous across layers and proposes layerwise learning rates based on weight spectrum shape, yielding up to 1.5×…
- Embodied AI 101
SimToolReal: Procedural Tool Generation and a Universal Objective for Zero-Shot Tool Manipulation
Trains generalist policies in simulation on procedurally generated tools to move objects, enabling real-world tool use across varied shapes/sizes.
- Embodied AI 101
Robometer and the Future of Robotic Reward Modeling
New framework for scalable robotic reward modeling using trajectory comparisons to train general-purpose reward models.
- Embodied AI 101
Qwen-VLA: A Generalist Vision–Language–Action Robot Model
A single generalist VLA built on Qwen3.5-4B + 1.15B DiT flow-matching action decoder that unifies manipulation, navigation, and trajectory prediction across 11…
- Embodied AI 101
EXPO-FT: Sample-Efficient Reinforcement Learning Fine-Tuning for Vision-Language-Action Models
Extends the EXPO method with real-world RL post-training for VLAs using image observations, action chunking, DAgger, and on-the-fly Q-value maximization…
- Embodied AI 101
RoboMeter: Learning Dense Rewards from Successes and Failures
RoboMeter trains dense reward models from both successful and failed robot trajectories, solving a key gap in prior methods that only learn from expert demos.
- Embodied AI 101
MobileGym: A Controllable, Parallel Sandbox for Mobile GUI Agents
Browser-hosted mobile environment with JSON state, deterministic judges, and 256 parallel rollouts. Reports +40.7 real-device points after GRPO training on 416…
- Embodied AI 101
ANY2ANY: Efficient Cross-Embodiment Transfer for Humanoid Whole-Body Tracking
Introduces a method to transfer a Unitree G1 foundation policy (Gear-Sonic) to LimX Oli/Luna humanoids using only 1% of the original compute/data. Achieves…
- Embodied AI 101
TriSplat: Feed-Forward 3D Reconstruction with Triangulated Meshes
Outputs physics-engine-compatible triangle meshes directly from sparse, unposed images without Gaussian splatting or post-processing.
- Embodied AI 101
MIKASA-Robo-VLA: A Memory-Intensive Benchmark for Vision-Language-Action Robotics
Releases a benchmark suite for systematically evaluating memory in Vision-Language-Action policies on tabletop manipulation tasks.
- Embodied AI 101
PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation
Introduces large-scale 3D world models pretrained on diverse real-world video to enable robust robotic manipulation policies that generalize beyond simulation.
- Embodied AI 101
Bimanual Pegboard Manipulation: A Benchmark for Vision-Language-Action Models
New LeRobot-based bimanual pegboard manipulation dataset with 52 episodes, 30k frames, 3 camera views, and 14-DOF arms for VLA evaluation. Provides…
- Embodied AI 101
FutureSim: Replaying Real-World Events to Evaluate AI Forecasting Agents
A benchmark designed to test AI models' capabilities in making accurate 3-month future predictions.
- Embodied AI 101
AgentFloor: A Benchmark for Long-Horizon Agent Planning
A 30-task benchmark for evaluating long-horizon planning capabilities across 16 different AI models.
- Embodied AI 101
AlexNet: The Deep Convolutional Network That Transformed Vision
AlexNet paper that sparked the modern deep learning revolution through convolutional neural networks.
- Embodied AI 101
A Few Useful Things to Know About Machine Learning
Practical insights into ML pitfalls and best practices for machine learning practitioners.
- Embodied AI 101
SimToolReal: A Universal Dexterous Tool-Use Policy
Introduces an object-centric sim-to-real policy that enables zero-shot dexterous tool use on physical robots without task-specific fine-tuning. Leverages…
- Embodied AI 101
Mimic-Video: Learning Physics Priors from Web-Scale Video for Robot Dexterity
Pretrains robot policies on large-scale web video to acquire dynamics and physics understanding instead of static images or VLMs. Yields faster training…
- Embodied AI 101
Deep Residual Learning for Image Recognition (ResNet)
Introduced residual connections (ResNet) enabling training of very deep networks, still widely used in modern architectures.
- Embodied AI 101
Attention Is All You Need – The Transformer Revolution
Introduced the Transformer architecture based purely on attention mechanisms, becoming the foundation of nearly all modern large language models.
- Embodied AI 101
NVIDIA Cosmos: World Foundation Models for Physical AI
World foundation models for video and physics prediction with SynthID watermarking for responsible AI practices. Developed in collaboration with Google…
- Embodied AI 101
LATENT: Teaching a Humanoid to Play Tennis from Imperfect Data
Introduces a three-stage pipeline that extracts a latent action space from noisy, low-quality human motion capture, then trains a high-level RL policy in…
- Embodied AI 101
CollabVR: Collaborative Video Reasoning with Vision-Language and Video Generation Models
Closed-loop framework coupling Vision-Language Models with Video Generation Models at step-level granularity. Mitigates long-horizon drift and mid-clip errors…
- Embodied AI 101
World Action Models: The Next Frontier in Embodied AI
First systematic survey defining World Action Models (WAMs) as embodied foundation models that jointly predict future states and generate actions. Covers…
- Embodied AI 101
Training a Whole-Body Control Foundation Model
Describes end-to-end learning of a foundation model for adaptive whole-body humanoid control via massive simulation variation. Combines proprioceptive…
- Embodied AI 101
DexJoCo: A Unified Benchmark for Task-Oriented Dexterous Manipulation
Releases an open-source MuJoCo-based benchmark with 11 dexterous tasks, low-cost teleoperation hardware, and 1.1K human demonstrations. Designed to evaluate…
- Embodied AI 101
MMSkills: Building Multimodal Skill Libraries for Visual Agents
Skill library, demonstrations, and dataset for multi-modal robotic skill learning and manipulation tasks.
- Embodied AI 101
PhysBrain 1.0 VLA (TwinBrainVLA): Dual-Brain Vision-Language-Action with Physics-Grounded Learning
Introduces dual-brain fusion Vision-Language-Action model with LangForce physics-grounded training methodology.
- Embodied AI 101
MolmoAct2-LIBERO: An Open Vision-Language-Action Model for Robotics
Vision-Language-Action (VLA) model fine-tuned on the merged LIBERO robotics dataset (1,693 episodes, 273k+ frames) achieving 98.25% success rate on…
- Embodied AI 101
SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Diffusion Transformers
A 2.6B-parameter open-source world model that generates coherent 720p, minute-long videos with precise 6-DoF camera control on a single GPU using a Hybrid…
- Embodied AI 101
WildClawBench: A Real-World, Long-Horizon Benchmark for AI Agents
New benchmark and dataset for robotic manipulation in unconstrained 'wild' environments. Includes standardized containers, leaderboards, and evaluation…
- Embodied AI 101
MCP-Cosmos: Bring Your Own World Model
Introduces a latent-space world model framework that lets agents simulate state transitions and iteratively refine plans before real-world execution. Evaluated…
- Embodied AI 101
OpenAI o1: Teaching LLMs to Think Slow and Deep
Details OpenAI's reasoning-focused o1 model and its 'long thought' approach using test-time compute scaling. Explores how extended reasoning during inference…
- Embodied AI 101
The Llama 3 Herd of Models
Comprehensive technical report on the Llama 3 family, covering architecture, training at scale, multimodal extensions, and real-world impact. Details the…
- Embodied AI 101
LATENT: Learning Athletic Humanoid Tennis Skills from Imperfect Human Motion Data
Introduces a three-stage pipeline that extracts a latent action space from low-quality human tennis demonstrations, then trains a high-level policy in…
- Embodied AI 101
AnyFlow: Any-Step Video Diffusion for Predictive World Modeling
First any-step video diffusion framework using flow maps, allowing a single model to adapt to arbitrary inference budgets for scalable high-quality video…
- Embodied AI 101
# Robotics: The Endgame
Technical roadmap mirroring LLM scaling: critiques VLAs, advocates video world models as second pretraining phase, introduces World Action Models (WAM)…
- Embodied AI 101
Claw-Eval: Toward Trustworthy and Transparent Evaluation of Autonomous Agents
Benchmark with 2,159 rubric items across 300 tasks using trajectory-aware grading and 3-trial Pass^3 scoring to mitigate luck. Evaluates agent reliability in…
- Embodied AI 101
LIBERO-Para: Paraphrase Robustness in Robotic Manipulation
Reveals paraphrase fragility in VLAs causing 22-52% success drops due to task misidentification. Introduces PRIDE metric weighting success by paraphrase…
- Embodied AI 101
YOR: Your Own Mobile Manipulator for Generalizable Robotics
Low-cost mobile manipulator design and training strategies for broad generalization in real-world tasks.
- Embodied AI 101
EgoSim: Egocentric World Simulator for Embodied Interaction Generation
Closed-loop egocentric video simulator maintaining persistent 3D scene state for consistent interactions, enabling cross-embodiment transfer from human videos…
- Embodied AI 101
Accelerating Video World Models: From Generative Videos to Real-Time Simulators
Comprehensive survey taxonomizing efficient architectures/algorithms for video world models as simulators, targeting compute bottlenecks in embodied AI…
- Embodied AI 101
From Tokens to Thoughts: Continuous Latent Reasoning in Large Models and Robot Control
Curated collection of 100+ works surveying shift to continuous latent spaces in LLMs/VLMs/VLAs for improved reasoning over discrete tokens, with relevance to…
- Embodied AI 101
CaP-X: Coding Agents for Physical eXecution
CaP-X is an open-source agentic robotics framework where LLMs/VLMs generate code to call perception and control APIs for execution across diverse simulated and…
- Embodied AI 101
DoRA: Weight-Decomposed Low-Rank Adaptation
An upgrade over LoRA for parameter-efficient fine-tuning, enabling better performance in LLMs by decomposing weights into magnitude and direction components.
- Embodied AI 101
AI Model Collapse: What Happens When AI Trains on Its Own Outputs
Seminal work showing how training on AI-generated data leads to 'model collapse' in neural networks, with urgent implications for future scaling.
- Embodied AI 101
PhAIL: Benchmarking Vision-Language-Action Models on Real-World Bin-Picking
Real-world hardware evaluation of VLAs on blind bin-to-bin picking, achieving max 64 picks/hour across hundreds of runs, with full videos/data exposing gaps in…
- Embodied AI 101
Co-training Large Behavior Models: Data Modalities and Training Strategies for Robot Manipulation
Comprehensive evaluation of 89 policies showing optimal co-training practices mixing real robot data with sim/egocentric human videos to boost diversity and…
- Embodied AI 101
HyDRA: Hybrid Memory for Dynamic Video World Models
Novel memory system preserving dynamic object identity and motion continuity across occlusions in video world models, addressing frozen/vanishing issues for…
- Embodied AI 101
# WildWorld: Dynamic World Modeling with Actions and Explicit State
Massive dataset enabling dynamic world models with explicit states and actions, supporting predictive modeling for cross-embodiment robotic control.
- Embodied AI 101
Omni-WorldBench: Evaluating Interactive 4D World Models
New benchmark assessing world models on interaction tasks, pushing predictive physics and video modeling towards robotics applications with action-conditioned…
- Embodied AI 101
SIMART: From Static Meshes to Sim-Ready Articulated Models
Unified MLLM framework with Sparse 3D VQ-VAE (70% token reduction) for part-level mesh decomposition and kinematic chain prediction, enabling physics-based…
- Embodied AI 101
EgoSim: An Egocentric World Simulator for Embodied Interaction
Closed-loop egocentric simulator persistently updating 3D scene state to generate spatially consistent interaction videos for continuous simulation, enabling…
- Embodied AI 101
Digit's New Motor Cortex: Sim-to-Real RL for Whole-Body Control
AI-trained capabilities for new whole-body motions using mocap/teleop data and sim-to-real reinforcement learning, deployable overnight on hardware.
- Embodied AI 101
EgoNav: Diffusion-Based Humanoid Navigation from Human Egocentric Video
Diffusion-based humanoid navigation trained solely on 5 hours of human egocentric video data, enabling zero-shot deployment on Unitree G1 for complex behaviors…
- Embodied AI 101
CaP-X: A Code-as-Policy Framework for Robot Manipulation
Comprehensive open-source agentic robotics framework treating VLMs/LLMs as code-generating APIs for perception (SAM3, Molmo) and control (IK, grasping), with…
- Embodied AI 101
Embodied Intelligence Breakthrough: Generalist AI’s GEN-1 Robots
We've created GEN-1, our latest milestone in scaling robot learning. We believe it to be the first general-purpose AI model that crosses a new performance…
- Embodied AI 101
CaP-X: LMs' First Physical Exam
A novel benchmark that evaluates language models on physical examination tasks, testing their ability to understand and perform clinical physical exam…
- Embodied AI 101
AI Model Collapse: The Danger of Training on AI-Generated Data
Demonstrated that LLMs trained recursively on AI-generated data suffer model collapse, a degenerative process where they lose grasp of true data distributions…