Barycentric Quantile Map Overview

• Barycentric quantile maps are constructions that use a common scalar parameter combined with transport and averaging to represent complex multivariate distributions.
• They leverage methods such as Brenier transport and HiMAP quantile maps to connect quantile representations with Wasserstein geometry and barycenter estimation.
• These approaches facilitate precise credible set formation, barycenter computation, and Fréchet regression, offering versatile tools for statistical and Bayesian inference.

Searching arXiv for papers directly relevant to barycentric quantile maps, transport-based quantiles, and barycentric projection. A barycentric quantile map is not a uniformly standardized object across the recent arXiv literature. The term is used most naturally for constructions that combine a common probability-indexing device with barycentric averaging, but the underlying mathematical realizations differ substantially. In one line of work, the closest object is a conditional vector quantile map defined as a Brenier transport from a fixed reference law to a posterior distribution, with nested images of latent balls yielding credible regions (Kim et al., 2024). In another, the closest object is a convex combination of optimal transport maps that generates Wasserstein barycenters under compatibility assumptions (Werenski et al., 2022). A third line introduces an explicit multivariate quantile-like representation, the HiMAP quantile map $\mathbf Q_\mu:[0,1]\to\mathbb R^d$, whose pointwise affine averages define well-posed multivariate barycenters and Fréchet regression estimators (Wang et al., 4 Mar 2026). Related work on barycentric projections, barycentric algebras, and contractive barycentric maps supplies adjacent geometric, variational, and algebraic interpretations, but does not itself define the term in a probabilistic quantile sense (You, 6 Jun 2026, Zamojska-Dzienio, 1 Jan 2025, Hiai et al., 2018).

1. Terminological status and conceptual scope

The recent literature does not present a single canonical definition of “barycentric quantile map.” Both "Deep Generative Quantile Bayes" and "Measure Estimation in the Barycentric Coding Model" explicitly state that they do not define an object with that name (Kim et al., 2024, Werenski et al., 2022). This suggests that the phrase is best treated as an umbrella term for several nearby constructions rather than as an established primitive.

Across these papers, three recurrent ideas organize the topic. First, a quantile-like parameterization provides a common coordinate or rank variable, such as a latent $u\in\mathcal U$ or a scalar $t\in[0,1]$. Second, a transport or pushforward map sends that parameter into the target distribution or target measure family. Third, a barycentric operation combines maps, displacement fields, or quantile representations across distributions. The precise meaning of “barycentric” therefore depends on context: it may refer to Wasserstein barycenters, convex combinations of Monge maps, or affine averaging in an induced function space.

A useful distinction is between transport-defined multivariate quantiles and barycentric averaging formulas. The former emphasize Brenier or Monge structure, as in conditional posterior simulation (Kim et al., 2024). The latter emphasize the algebra of averaging maps or displacement vectors, as in the barycentric coding model and HiMAP (Werenski et al., 2022, Wang et al., 4 Mar 2026). A plausible implication is that the phrase “barycentric quantile map” is most precise when a single indexing variable has a distribution-invariant meaning and supports pointwise averaging across measures.

2. Optimal-transport quantiles and conditional Brenier maps

In the most transport-theoretic usage, the closest analogue to a barycentric quantile map is the vector quantile map $Q_P$, defined for a target law $P$ on $\mathbb R^d$ as the gradient of a convex potential $\psi$. When $P$ has finite second moments, $Q_P$ is exactly the Monge map minimizing quadratic cost,

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$

and by Brenier’s theorem one has $u\in\mathcal U$0 (Kim et al., 2024). This is the paper’s main bridge between quantiles and Wasserstein geometry.

For posterior simulation, the same paper uses the conditional version

$u\in\mathcal U$1

with the defining pushforward relation

$u\in\mathcal U$2

The base distribution $u\in\mathcal U$3 is supported on the unit Euclidean ball $u\in\mathcal U$4, with $u\in\mathcal U$5, $u\in\mathcal U$6, $u\in\mathcal U$7 uniform on the sphere $u\in\mathcal U$8, and $u\in\mathcal U$9 (Kim et al., 2024). In this framework, a latent “rank” $t\in[0,1]$0 is mapped to a posterior draw $t\in[0,1]$1, and the image of the reference law under the map is the posterior itself.

The same paper also gives the inverse conditional rank map

$t\in[0,1]$2

where $t\in[0,1]$3 is the convex conjugate in the $t\in[0,1]$4-variable. This duality is central because it identifies quantile structure with OT geometry rather than with a Wasserstein-barycenter construction. Accordingly, the paper’s own characterization is that its geometry is OT/Brenier/MK-depth geometry rather than an explicit Wasserstein-barycenter construction (Kim et al., 2024).

A related but distinct plan-to-map construction appears in "Barycentric Projections of Optimal Transport Plans on Riemannian Manifolds." There the intrinsic barycentric projection assigns to each source point $t\in[0,1]$5 the conditional Fréchet mean of its conditional destination law,

$t\in[0,1]$6

and is shown to be the best deterministic representative of a coupling under squared geodesic loss (You, 6 Jun 2026). This is not a quantile map in the classical sense, but it is a canonical barycentric surrogate of a transport object. In Euclidean space it reduces to conditional expectation, while in Monge cases it recovers the actual transport map (You, 6 Jun 2026). This suggests a broader interpretation in which barycentric quantile maps can also arise as deterministic representatives extracted from probabilistic transport couplings.

3. Barycentric transport maps in Wasserstein barycenter geometry

"Measure Estimation in the Barycentric Coding Model" develops the barycentric side of the topic. It studies an unknown measure $t\in[0,1]$7 assumed to lie in the set of Wasserstein-2 barycenters generated by known reference measures $t\in[0,1]$8, with barycenter

$t\in[0,1]$9

where the $Q_P$0 are barycentric coordinates (Werenski et al., 2022).

The paper does not introduce quantile functions, but it identifies the nearest object to a barycentric quantile map as a barycentric combination of optimal transport maps. Let $Q_P$1 denote the optimal map from $Q_P$2 to $Q_P$3. The gradient of the variance functional is

$Q_P$4

so the barycenter condition is expressed as a balancing relation among displacement maps in the Wasserstein tangent space (Werenski et al., 2022). The associated Gram matrix

$Q_P$5

reduces coordinate recovery to the convex quadratic program

$Q_P$6

The strongest map formula appears under the paper’s compatibility assumption, where for all $Q_P$7,

$Q_P$8

In that regime the barycenter admits the exact pushforward representation

$Q_P$9

Equivalently,

$P$0

This is the paper’s closest explicit realization of a barycentric map (Werenski et al., 2022).

The paper notes that continuous measures on $P$1 are among the compatible classes, but it does not state the 1D formula

$P$2

That quantile representation is described only as an inference by standard OT theory, not as a theorem of the paper (Werenski et al., 2022). This distinction is important: the paper proves a barycentric transport-map characterization, not a quantile-function formalism.

4. HiMAP and explicit multivariate barycentric quantile maps

"HiMAP: Hilbert Mass-Aligned Parameterization for Multivariate Barycenters and Frećhet Regression" provides the most direct realization of a barycentric quantile map. For a multivariate probability measure $P$3, the paper defines the HiMAP quantile map

$P$4

designed so that the scalar parameter $P$5 has a common, distribution-invariant “mass level” meaning across different measures (Wang et al., 4 Mar 2026).

The construction recursively partitions the support through equiprobable conditional-median splits arranged according to Hilbert curve ordering rules. Starting from a bounded axis-aligned box $P$6, the split at depth $P$7 is the conditional median

$P$8

so that each child cell has half the mass of its parent (Wang et al., 4 Mar 2026). This makes the recursion mass aligned rather than geometrically uniform.

For fixed depth $P$9, the finite-depth representative is

$\mathbb R^d$0

Under the assumptions that $\mathbb R^d$1 admits a density $\mathbb R^d$2 on $\mathbb R^d$3 with $\mathbb R^d$4, conditional medians are unique at every depth, and the splitting schedule is balanced, the paper proves geometric decay of cell diameters,

$\mathbb R^d$5

and therefore existence of the limit

$\mathbb R^d$6

A crucial theorem states

$\mathbb R^d$7

so $\mathbb R^d$8 is a measurable map from a common scalar mass parameter to the target distribution (Wang et al., 4 Mar 2026).

The representation induces the discrepancy

$\mathbb R^d$9

and for $\psi$0 this is the Hilbert norm

$\psi$1

The defining structural property is closure under affine averaging: for any weights satisfying $\psi$2,

$\psi$3

is itself the HiMAP quantile map of a valid probability measure (Wang et al., 4 Mar 2026). This yields the explicit barycenter formula

$\psi$4

with

$\psi$5

Within the cited literature, this is the clearest exact meaning of barycentric quantile map (Wang et al., 4 Mar 2026).

The same closure property makes HiMAP especially suitable for Fréchet regression with affine weights that sum to one but may be negative. The paper gives the population regression identity

$\psi$6

and the empirical estimator

$\psi$7

(Wang et al., 4 Mar 2026). This generalizes the 1D algebra of quantile averaging to multivariate responses in an induced Hilbert geometry rather than in exact $\psi$8 geometry.

5. Bayesian posterior geometry and nested credible regions

In "Deep Generative Quantile Bayes," the conditional vector quantile map is used as a posterior generator rather than as a barycenter operator. Training data consist of simulated triples

$\psi$9

where $P$0, $P$1, and $P$2 i.i.d. (Kim et al., 2024). The learned map has domain $P$3, conditioning variable $P$4, codomain $P$5, and pushforward

$P$6

A core modeling step is the assumption of a learned sufficient summary statistic $P$7 satisfying

$P$8

together with the affine-in-summary potential parameterization

$P$9

The neural architecture uses ICNNs for $Q_P$0 and $Q_P$1, with “3 hidden layers of width 512, and CELU activation,” and combines a DeepSet $Q_P$2 with an LSTM $Q_P$3 to define

$Q_P$4

(Kim et al., 2024). The paper’s central optimization problem is the empirical dual OT loss

$Q_P$5

The paper’s most geometric contribution is its use of Monge–Kantorovich depth for credible sets. For an unconditional distribution $Q_P$6, the MK-depth region of content $Q_P$7 is

$Q_P$8

where $Q_P$9 is the ball of radius $$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$0, yielding nested regions

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$1

In the posterior setting this becomes

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$2

Thus posterior credible sets are direct images of concentric latent balls under the conditional Brenier map (Kim et al., 2024).

The paper proves uniform consistency of the estimated potential and conjugate, consistency of the vector quantile map,

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$3

consistency of the recovered posterior,

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$4

and Hausdorff convergence of credible sets,

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$5

(Kim et al., 2024). The Gaussian conjugate experiment also displays support shrinkage, meaning that posterior contours contract as sample size $$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$6 increases. In this setting, the “barycentric” appearance comes from centered latent balls being transported into posterior contours, but the paper is explicit that this is not a Wasserstein-barycenter construction (Kim et al., 2024).

6. Related abstractions, limitations, and common misconceptions

Several nearby theories clarify what a barycentric quantile map is not. "Partitions of Unity and Barycentric Algebras" studies convex combinations on polytopes through barycentric algebras and the tautological map

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$7

from partitions of unity to points in a convex polytope (Zamojska-Dzienio, 1 Jan 2025). This is a canonical barycentric coordinate map, but the paper explicitly does not define a probabilistic quantile map. Its relevance is algebraic: it shows how weight vectors or simplex-valued functions can be converted into points by barycentric combination.

"Convergence theorems for barycentric maps" studies a different abstraction: a contractive barycentric map

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$8

satisfying

$$Q_P = \arg\min_{Q:\,Q\#F_U=P}\; \mathbb E_{U\sim F_U}\|Q(U)-U\|^2,$$9

and defines nonlinear conditional expectation by

$u\in\mathcal U$00

(Hiai et al., 2018). This again supplies a barycentric summary of measures, together with martingale, ergodic, continuity, and large deviation theory, but it does not define quantile maps or transport-based rank parameterizations.

Several misconceptions therefore need to be separated.

Misconception	Correction
A barycentric quantile map is a standardized OT term	The cited papers do not present a single standardized definition
It always means a Wasserstein barycenter formula	In (Kim et al., 2024) the relevant object is a Brenier quantile transport, not a barycenter
It is always an exact multivariate analogue of 1D quantiles	HiMAP provides such an analogue only in an induced HiMAP geometry, not in exact $u\in\mathcal U$01 geometry (Wang et al., 4 Mar 2026)
Any barycentric projection is a quantile map	Barycentric projection is a deterministic surrogate of a coupling, not a classical quantile function (You, 6 Jun 2026)

The most accurate synthesis is therefore plural. In OT-based Bayesian computation, the nearest object is the conditional vector quantile map $u\in\mathcal U$02 (Kim et al., 2024). In Wasserstein barycenter geometry, the nearest object is the convex combination of optimal transport maps $u\in\mathcal U$03 under compatibility (Werenski et al., 2022). In multivariate quantile-like barycenter calculus, the most explicit object is the HiMAP quantile map $u\in\mathcal U$04, whose pointwise affine averages define barycenters and Fréchet regression estimators (Wang et al., 4 Mar 2026). Taken together, these works indicate that “barycentric quantile map” is best understood as a family of constructions linking common mass coordinates, transport maps, and barycentric averaging, with the exact meaning determined by the ambient geometry and the target problem.