Re­search to­ward a the­ory of ab­strac­tion suit­able for em­bed­ded agency.

Key back­ground con­cepts:

Both of these will be lev­er­aged very heav­ily in this se­quence, so it’s worth check­ing them out be­fore div­ing in.

This work is sup­ported by a grant from the long-term fu­ture fund.

What is Ab­strac­tion?

Causal Ab­strac­tion Toy Model: Med­i­cal Sensor

Ex­am­ples of Causal Abstraction

Ab­strac­tion, Causal­ity, and Embed­ded Maps: Here Be Monsters

Causal Ab­strac­tion Intro

Defi­ni­tions of Causal Ab­strac­tion: Re­view­ing Beck­ers & Halpern

How to Throw Away In­for­ma­tion in Causal DAGs

Ex­am­ple: Markov Chain

Log­i­cal Rep­re­sen­ta­tion of Causal Models

(A → B) → A in Causal DAGs

For­mu­lat­ing Re­duc­tive Agency in Causal Models

Trace: Goals and Principles


Ab­strac­tion = In­for­ma­tion at a Distance

Me­di­a­tion From a Distance

Noise Simplifies

In­te­grat­ing Hid­den Vari­ables Im­proves Approximation

In­tu­itions on Univer­sal Be­hav­ior of In­for­ma­tion at a Distance

Mo­ti­vat­ing Ab­strac­tion-First De­ci­sion Theory

Writ­ing Causal Models Like We Write Programs

Point­ing to a Flower

Public Static: What is Ab­strac­tion?

Carte­sian Boundary as Ab­strac­tion Boundary

Causal­ity Adds Up to Normality

The In­dex­ing Problem

Ab­strac­tion, Evolu­tion and Gears