EngineersofAI LogoCoursesBlogResearch LabAI LettersThe LabCode BankInteractive 3DKodrEarnestJobs
Skip to main content
Interactive 3D/Causal DAG Explorer
UXZYW
Hover nodes for descriptions ยท Click to select
Presets
Conditioning
Selected Node
Click a node to see details
Key Concepts
Confounder (U): causes both X and Y. Creates false correlation.
Mediator (Z): on the causal path. Blocking it removes the effect.
Collider (W): two arrows in. Conditioning on it opens spurious paths.

Causal DAG Explorer - Interactive Visualization

Causal inference requires more than correlation - it requires understanding the causal graph. Confounders (common causes) create spurious correlations. Mediators (on the causal path) should not be controlled for. Colliders (both caused by X and Y) open spurious paths when conditioned on. This SVG-based explorer lets you build DAGs, identify these structures, and check d-separation between node pairs.

  • Build DAGs by clicking to add directed edges
  • Identify confounders, mediators, and colliders with presets
  • Check d-separation between any two nodes
  • See how conditioning on a node opens/closes backdoor paths
  • Foundation for causal ML, A/B test design, and debiasing models

Part of the EngineersOfAI Interactive 3D - free interactive visualizations covering every major concept in machine learning and AI engineering. Hover any element for a plain-English explanation. No code required.