ATLAS / A SPATIAL KNOWLEDGE INDEX

How the ideas connect.

Choose a concept. Follow its relationships.
27 CONCEPTS / 39 TYPED RELATIONS

ATLAS / LAYERED KNOWLEDGE MAP

A stable spatial index for concepts, their evidence, and the bridges between domains.

SCROLL HORIZONTALLY TO EXPLORE ALL LAYERS ↔

01 / FOUNDATIONS
02 / ARTIFICIAL INTELLIGENCE
03 / ROBOTICS
04 / EMBODIED INTELLIGENCE
MAMathematics &StatisticsPRProgramming &SoftwareCSComputer ScienceEMElectronics &Embedded SystemsMLMachine LearningDLDeep LearningCVComputer VisionLMLanguage Models &LLMsVMMultimodal Learning& VLMsRLReinforcementLearningAGAI Agents & AgenticSystemsRCRobot Mechanics,Kinematics &ControlRPSensing & RobotPerceptionNVSLAM & NavigationRSRobot SoftwareSystems &CommunicationMGManipulation &GraspingLHLegged & HumanoidRoboticsSIMRobotics Simulation& ToolchainsEPEmbodied Perception& SpatialIntelligenceENEmbodied NavigationEDEmbodiedManipulation &DexterityHBHumanoid &Whole-BodyIntelligenceVLARobot FoundationModels & VLAWMWorld Models &Embodied MemoryEAEmbodied Agents &Long-HorizonAutonomyRLNRobot Learning &Policy LearningDSEmbodied Data,Simulation &Sim-to-Real
prerequisitefoundation / enablesapplied inintegrationSELECT TO INSPECT · DRAG ON DESKTOP
COMPLETE INDEX / NO SCRIPT REQUIRED

Every concept, still connected.

Use the index when you want the full semantic record rather than the bounded visual map. 按技术演进层次快速检索底层理论、学习感知、物理控制与具身闭环核心节点。

01

Foundations

怎么表示、计算和实现?

数学 · 编程 · CS · 嵌入式

02

Artificial Intelligence

怎么学习、感知、推理?

ML · DL · CV · LLM · VLM · RL · Agents

03

Robotics

怎么让真实机器人运动起来?

Mechanics · Perception · SLAM · ROS 2 · Manipulation · Humanoids

04

Embodied Intelligence

怎么让智能体在真实世界完成任务?

Perception · Navigation · VLA · World Models · Robot Learning

Prerequisites
Dark dashed lines point to the ideas a node assumes you already understand.

Typed relations
Each line records a prerequisite, foundation, application, integration, or frontier bridge.

Connected work
Every node can open its canonical record with grouped Blog, Papers, and Projects links.

Four layers
Foundations, Artificial Intelligence, Robotics, and Embodied Intelligence form the core taxonomy.

Embodied intelligence
The fourth layer groups capabilities, models, cognition, learning, and deployment.

Progressive disclosure
The complete semantic index remains below the bounded visual map.