nLab observational monad

Contents

Context

Measure and probability theory

Higher algebra

Monoidal categories

monoidal categories

With braiding

With duals for objects

With duals for morphisms

With traces

Closed structure

Special sorts of products

Semisimplicity

Morphisms

Internal monoids

Examples

Theorems

In higher category theory

Contents

Idea

An observational monad is a monad where, intuitively, one can distinguish two monadic (for example probabilistic) values by means of taking repeated independent observations.

It is of particular importance in categorical probability, where these observations correspond to iid random variables.

Definition

Let CC be a category with products, and let PP be a symmetric monoidal monad (or commutative monad) on CC.

Denote also by C TC_T its Kleisli category, which is then a copy-discard category. As usual, we have an adjunction L:CC T,R:C TCL:C\to C_T, R:C_T\to C such that P=RLP=R\circ L.

Denote the comonad LRL\circ R on C TC_T again by PP. (Note that this equips C TC_T with a thunk-force structure.) Finally, denote ε X:PXX\varepsilon_X:P X\to X the components of the counit of the adjunction. (In categorical probability, this is known as the sampling map.)

The monad PP is called observational if and only if for every object XX, the family of morphisms of C TC_T

{PXcopy(PX) nε nX n,n} \Big\{ P X \xrightarrow{\quad\mathrm{copy}\quad} (P X)^{\otimes n} \xrightarrow{\quad\varepsilon^{\otimes n}\quad} X^{\otimes n}\;,\; n \in \mathbb{N} \Big\}

is jointly monic.

The intuition is that one can distinguish two morphisms (for example, states) into PXP X by sampling independently as many times as needed.

A representable Markov category whose probability monad is observational is called an observationally representable Markov category.

Properties

  • The Giry monad is observational thanks to the pi-lambda theorem? or, equivalently, the monotone class theorem?. (Moss-Perrone’22)

  • The Kleisli category of an observational monad, with its canonical copy-discard and thunk-force structures, is such that every deterministic morphism is thunkable (the converse is always true). (Moss-Perrone’22)

  • Every observationally representable Markov category is necessarily almost-surely-compatibly representable. That is, in the bijection

    f:XYf :X(det)PY f:X\xrightarrow{\qquad} Y \qquad\leftrightarrow\qquad f^\sharp:X \xrightarrow{\;(det)\;} P Y

    we have that for every state (probability measure) pp on XX, two morphisms f,g:XYf,g:X\to Y are pp-almost surely equal? if and only if f f^\sharp and g g^\sharp are. (FGLPS)

  • Observationality is related to de Finetti's theorem: stating the theorem as a limit, observationality is closely related to the uniqueness part in the universal property. It coincides exactly with this uniqueness whenever the necessary Kolmogorov products exist. (Moss-Perrone’22)

See also

References:

The concept, in the form presented here, was first introduced in:

It also appears in:

  • Tobias Fritz, Tomáš Gonda, Antonio Lorenzin, Paolo Perrone and Dario Stein, Absolute continuity, supports and idempotent splitting in categorical probability, 2023. (arXiv)

  • Noé Ensarguet, Paolo Perrone, Categorical probability spaces, ergodic decompositions, and transitions to equilibrium, arXiv:2310.04267

  • Tobias Fritz and Antonio Lorenzin, Involutive Markov categories and the quantum de Finetti theorem, 2023. (arXiv)

Related papers:

category: probability

Created on September 28, 2024 at 10:15:55. See the history of this page for a list of all contributions to it.