Causal Emergence: When Distortions In A Map Obscure The Territory

on YouTube ↗38m

" summary

Argues that causal emergence is defined by finding a coarsening partition of a finite-state transition matrix that maximizes effective information, computed as KL divergence / mutual information under maximum entropy interventions, and searches over all coarse-grainings of I and J. Shows simple 8-state examples where seven microstates collapse despite differing effects (sensitive to changes like 0.7→0.8), proves abstraction and marginalization do not commute, and concludes the channel‑capacity analogy and uniqueness of emergent partitions are dubious.

AI-written summary · about

" chapters (6)

1. 0:00Introduction
2. 9:09Theory
3. 11:48Effective Information
4. 16:15Causal Emergence
5. 22:35Example
6. 25:10What bothers me

" related talks

• 48mCausal Discovery in the Quantum ContextSally Shrapnel · Quantum Physics and Statistical Causal Models (Simons Institute workshop, 2022)
• 37mAlgorithmic Fairness From The Lens Of Causality And Information TheorySanghamitra Dutta · Algorithmic Aspects of Causal Inference 2022
• 24mA Friendly Introduction To CausalityAlex Dimakis · MLconf SF 2016
• 01h 06mIntroduction to Causal Graphical Models: Graphs, d-separation, do-calculusSpencer Gordon · Causality Boot Camp
• 48mSpatial Coupling as a Proof TechniqueNicolas Macris · Random Instances and Phase Transitions (Simons Institute, 2016)
• 43mInformation-Theoretic Bounds and Phase Transitions in Community Detection and High-Dimensional ClustCris Moore · Random Instances and Phase Transitions (Simons Institute, 2016)
• 01h 01mThe Mathematics of Causal Inference, with Reflections on Machine Learning and the Logic of ScienceJudea Pearl · Symposium on Visions of the Theory of Computing 2013
• 29mEfficient Discovery of Actual Causality in Stochastic SystemsArshia Rafieioskouei, Kenneth Rogale, Borzoo Bonakdarpour · VMCAI 2026