---
title: Genuine Reflected Multi-Entropy Overview
url: https://www.emergentmind.com/topics/genuine-reflected-multi-entropy
type: topic
---

# Genuine Reflected Multi-Entropy Overview

Genuine reflected multi-entropy denotes a still non-uniform class of multipartite constructions at the intersection of reflected entropy, reflected multi-entropy, and genuine multi-entropy. In the narrowest current usage, it refers to the quantity \(GS_{R(q-1)}\), introduced as a reflected-entropy-based object that is required to vanish on pure states containing only lower \(q'<q\)-partite entanglement [2507.15262]. In a broader historical sense, the same phrase points toward earlier multipartite reflected-entropy candidates such as \(\Delta_R\), candidate holographic \(n\)-party reflected entropies, and the later reflected multi-entropy \(S_R^{(q)}\), none of which by themselves were defined to isolate irreducible multipartite structure [1909.10456], [1909.03154], [2410.08546]. Any such program is constrained by the no-go result that ordinary reflected entropy and the Rényi reflected entropies \(S_R^{(\alpha)}\) for \(0<\alpha<2\) are not universal correlation measures under partial trace [2302.10208].

## 1. Terminological status and conceptual scope

The literature does not use a single standardized meaning for “genuine reflected multi-entropy.” Some papers explicitly introduce multipartite reflected quantities, others introduce genuine multi-entropy without reflection, and only one of the sources supplied here defines an object under the name “genuine reflected multi-entropy” in a direct prescription-based way [2507.15262]. This makes the term historically layered rather than settled.

The main nearby notions can be organized as follows.

| Object | Defining cue | Status relative to the term |
|---|---|---|
| Reflected multi-entropy \(S_R^{(q)}\) | Pure-state multi-entropy evaluated on canonical purification | Multipartite reflected, not genuine |
| Genuine reflected multi-entropy \(GS_{R(q-1)}\) | Linear combination of reflected multi-entropies and ordinary multi-entropies | Direct narrow usage |
| Multipartite reflected entropy \(\Delta_R\) | Symmetric generalized reflected entropy from iterated purification | Early intrinsic multipartite candidate |
| L-entropy \(\ell_{A_1\cdots A_n}\) | Geometric mean of pairwise reflected-entropy upper-bound deficits | Reflected-inspired, not reflected multi-entropy |

Within this landscape, “genuine” has two distinct meanings. In the subtraction-based multi-entropy literature, it means removal of all lower-partite contributions so that the quantity vanishes on states with only lower-order entanglement. In the older holographic reflected-entropy literature, it often means sensitivity to intrinsically multipartite wedge structure rather than a sum of bipartite terms. These meanings overlap conceptually but are not formally identical [2507.15262], [1909.03154].

## 2. Reflected multi-entropy as the immediate precursor

The direct precursor of genuine reflected multi-entropy is the reflected multi-entropy introduced as a mixed-state generalization of pure-state multi-entropy. For a mixed \(q\)-partite density matrix \(\rho_{A_1A_2\cdots A_q}\), one canonically purifies it and then evaluates the pure-state multi-entropy on the doubled subsystems. In the tripartite case,
\[
S_R^{(q=3)}(A;B;C)=S^{(q=3)}(AA^*;BB^*;CC^*)_{\ket{\sqrt{\rho_{ABC}}}},
\]
and more generally
\[
S_R^{(q)}(A;B;\cdots)=S^{(q)}(A_1A_1^*;A_2A_2^*;\cdots)_{\ket{\sqrt{\rho}}}.
\]
Its replica definition introduces two indices, \(m\) for the canonical-purification construction and \(n\) for the multipartite Rényi structure,
\[
S_R^{(q)}=\lim_{m\to 1}\lim_{n\to 1}\frac{1}{1-n}\log\frac{\mathcal Z_{n^{q-1},m}}{(\mathcal Z_{1,m})^{n^{q-1}}}.
\]
By construction it reduces to twice pure-state multi-entropy on pure states, to ordinary reflected entropy at \(q=2\), and it vanishes on fully factorized states [2410.08546].

This object is multipartite and reflected, but not genuine in the strict subtraction-based sense. The same source is explicit that it does not define a separate “genuine reflected multi-entropy,” does not subtract lower-order multipartite pieces, and does not prove that its tripartite member isolates only irreducible tripartite correlations. Its holographic dual is a multipartite minimal web,
\[
S_R^{(q)}=\frac{2}{4G_N}\min_{\mathcal W}L[\mathcal W],
\]
and the tripartite case was checked in AdS\(_3\)/CFT\(_2\) through a six-point twist correlator at large \(c\) [2410.08546].

Earlier holographic work had already sought multipartite reflected quantities by more geometric routes. One line constructed special generalized reflected entropies by repeated canonical purification and singled out a symmetric tripartite object \(\Delta_R(A:B:C)\), proposed to satisfy
\[
\Delta_R(A:B:C)=2\Delta_W(A:B:C),
\]
with \(\Delta_W\) the multipartite entanglement wedge cross section [1909.10456]. Another line proposed two candidate \(n\)-party reflected entropies obtained by replica gluing followed by canonical purification, with leading large-\(N\) relations of the form
\[
S_R(A_1:\ldots:A_n)=I(n)\,E_W(A_1:\ldots:A_n),
\]
where the normalization depends on the construction and on parity [1909.03154].

## 3. Prescription-based definition of genuine reflected multi-entropy

The explicit modern definition of genuine reflected multi-entropy is given in a program motivated by the failure of the ordinary multipartite Markov gap to isolate genuine \(q\)-partite entanglement. Starting from a pure \(q\)-partite state \(\ket{\psi}_{A_1A_2\cdots A_q}\), one traces out \(A_q\) and considers the \((q-1)\)-partite mixed state
\[
\rho_{A_1A_2\cdots A_{q-1}}=\mathrm{Tr}_{A_q}\!\left[\ket{\psi}\bra{\psi}\right].
\]
One then canonically purifies \((\rho_{A_1\cdots A_{q-1}})^m\) and defines the reflected multi-entropy
\[
S_{R(q-1)}^{(m,n)}(A_1:\cdots:A_{q-1})
=
S_{n(m)}^{(q-1)}(A_1A_1^*:\cdots:A_{q-1}A_{q-1}^*).
\]
The ordinary multipartite Markov gap is
\[
MG^{M(q-1)}(A_1:\cdots:A_{q-1})
=
S_{R(q-1)}(A_1:\cdots:A_{q-1})
-
\Big[\sum_{i=1}^{q-1}S(A_i)-S(A_1\cdots A_{q-1})\Big].
\]
The motivation for introducing \(GS_{R(q-1)}\) is that this ordinary multipartite Markov gap is not genuinely \(q\)-partite: for example, for a partially separable state such as \(\ket{\psi}_{ABCD}=\ket{\psi_1}_A\otimes\ket{\psi_3}_{BCD}\), the 4-party quantity need not vanish and reduces to residual lower-partite structure [2507.15262].

The genuine reflected multi-entropy is therefore defined operationally by prescriptions rather than by a closed universal formula:
\[
GS_{R(q-1)}(A_1:\dots:A_{q-1})\coloneqq
\lim_{m\to 1}\lim_{n\to 1}GS_{R(q-1)}^{(m,n)}(A_1:\dots:A_{q-1}).
\]
The construction is required to satisfy three conditions. First, it must vanish for all pure states that are only \(q'\)-partite entangled with \(q'<q\). Second, it must be a linear combination of reflected multi-entropies \(S_{R(q'-1)}^{(m,n)}\) with \(q'\le q\) and ordinary multi-entropies \(S_n^{(q')}\) with \(q'<q\), while treating \(A_1,\dots,A_{q-1}\) symmetrically and \(A_q\) as the distinguished traced-out subsystem. Third, at \((m,n)=(2,2)\) it must satisfy
\[
GS_{R(q-1)}^{(2,2)}(A_1:\dots:A_{q-1})=2\,\mathrm{GM}^{(q)}_2(A_1:\dots:A_q).
\]
In this form, genuine reflected multi-entropy is explicitly defined as a reflected-entropy-based analogue of genuine multi-entropy rather than as a direct multipartite canonical-purification entropy [2507.15262].

## 4. The explicit \(q=4\) construction

The first nontrivial explicit case is \(q=4\), where the traced-out subsystem is \(D\) and the remaining reflected quantity is tripartite. The compact formula given for the Rényi object is
\[
\begin{aligned}
GS_{R(3)}^{(m,n)}(A:B:C)
=\;& S_{R(3)}^{(m,n)}(A:B:C) \\
&-\frac12\Big[ S_{R(2)}^{(m,n)}(AB:C) +S_{R(2)}^{(m,n)}(AC:B) +S_{R(2)}^{(m,n)}(A:BC) \Big] \\
&+\frac13\Big( \mathrm{GM}_n^{(3)}(AB:C:D) +\mathrm{GM}_n^{(3)}(AC:B:D) +\mathrm{GM}_n^{(3)}(BC:A:D) \Big) \\
&-\frac23\Big( \mathrm{GM}_n^{(3)}(AD:B:C) +\mathrm{GM}_n^{(3)}(BD:A:C) +\mathrm{GM}_n^{(3)}(CD:A:B) \Big) \\
&+\left(\frac16-2a\right)I_n^{(3)}(A:B:C),
\end{aligned}
\]
with
\[
I_n^{(3)}(A:B:C)=S_n(A)+S_n(B)+S_n(C)-[S_n(AB)+S_n(AC)+S_n(BC)]+S_n(ABC),
\]
and
\[
GS_{R(3)}(A:B:C)=\lim_{m\to1}\lim_{n\to1}GS_{R(3)}^{(m,n)}(A:B:C).
\]
This formula is the paper’s central explicit realization of the concept [2507.15262].

Several properties are established for this \(q=4\) construction. It vanishes on any pure state on \(ABCD\) whose entanglement is only \(q'<4\)-partite. The paper checks this explicitly for \(\ket{\psi_3}_{ABC}\otimes\ket{\psi_1}_D\) and \(\ket{\psi_3}_{ABD}\otimes\ket{\psi_1}_C\), and then extends the result by additivity to arbitrary lower-partite tensor-product constructions such as \(\ket{\psi_2}_{AB}\otimes\ket{\psi_2}_{CD}\). The construction is additive under tensor products of pure states, symmetric under permutations of \(A,B,C\), and not fully symmetric in all four parties because \(D\) is distinguished as the traced-out subsystem. At \((m,n)=(2,2)\), it obeys the exact bridge relation
\[
GS_{R(3)}^{(2,2)}(A:B:C)=2\,\mathrm{GM}_2^{(4)}(A:B:C:D).
\]
For the four-party GHZ state, the paper finds
\[
GS_{R(3)}^{(m,n)}(A:B:C)=\left(\frac16-2a\right)\log 2,
\]
so choosing \(a=\frac{1}{12}\) makes the quantity vanish on \(\ket{\mathrm{GHZ}_4}\) [2507.15262].

The same source is equally explicit about what remains unproved. It does not provide a closed general-\(q\) formula, does not prove nonnegativity or monotonicity in full generality, does not supply an operational recovery theorem, and does not give a holographic dual for \(GS_R\). Thus the explicit \(q=4\) case is a constructional proof of principle rather than a finished general theory.

## 5. Relations to holography, replica invariants, and ordinary genuine multi-entropy

Genuine reflected multi-entropy sits within a broader network of replica constructions. On the reflected side, tripartite and \(n\)-partite reflected quantities had already been proposed holographically. The symmetric \(\Delta_R\) of iterated canonical purification was designed as a multipartite reflected entropy dual to \(2\Delta_W\), where \(\Delta_W\) was interpreted as an intrinsic three-body entanglement wedge cross section [1909.10456]. Separately, multipartite reflected entropy candidates built by replica gluing plus canonical purification were introduced as boundary von Neumann entropies dual to integer multiples of multipartite \(E_W\), explicitly to avoid collapsing into sums of bipartite reflected entropies [1909.03154]. These constructions supply an earlier holographic meaning for “genuine” in the sense of sensitivity to multiparty wedge structure, but they are not the same as the later subtraction-based \(GS_R\).

On the multi-entropy side, ordinary genuine multi-entropy \(GM_n^{(q)}\) is a distinct object built from inclusion-exclusion-style subtractions of ordinary multi-entropies. It appears directly in the defining condition
\[
GS_{R(q-1)}^{(2,2)}=2\,GM_2^{(q)},
\]
and therefore acts as the reference genuine measure to which the reflected construction is anchored [2507.15262]. This bridge is reinforced by a separate recursion result at Rényi-2:
\[
S_{R(q-1)}^{(m=2,n=2)}(A_1:\cdots:A_{q-1})
=
2\Big[S_2^{(q)}(A_1:\cdots:A_q)-S_2(A_1\cdots A_{q-1})\Big],
\]
which rewrites ordinary \(q\)-partite multi-entropy in terms of reflected multi-entropy of a canonically purified reduced state. For \(q=3\), this becomes
\[
S_R^{(2,2)}(A_1:A_2)=2\Big[S_2^{(3)}(A_1:A_2:A_3)-S_2(A_1A_2)\Big].
\]
This does not yet define a genuine reflected multi-entropy, but it makes reflected multi-entropy an algebraic component of the ordinary multi-entropy hierarchy at Rényi-2 [2602.16331].

Replica-invariant work on tripartite pure states draws a similar line between the “genuine” and “reflected” sectors. The genuine multi-entropy
\[
G_n^{(3)}(A:B:C)=S_n^{(3)}(A:B:C)-\frac12\big(S_n(A)+S_n(B)+S_n(C)\big)
\]
is treated there as the genuine quantity, while a separate dihedral invariant is proved to satisfy
\[
\mathcal D_{2n}(A:B)=S^R_{2,n}(A:C)
\]
for general tripartite pure states. The result clarifies that the reflected structure and the genuine subtraction structure are parallel but not identical [2509.00593].

## 6. No-go results, limitations, and adjacent programs

Any encyclopedic account of genuine reflected multi-entropy must include the strongest negative result in the background literature: reflected entropy is not, in general, a genuine measure of correlations. The minimal requirement for a bipartite correlation measure \(\delta_\rho(A:B)\) is monotonicity under discarding subsystems,
\[
\delta_\rho(A:BC)\ge \delta_\rho(A:B).
\]
For reflected entropy, explicit counterexamples show
\[
S_{R,\rho}(A:BC)<S_{R,\rho}(A:B),
\]
and, more strongly, none of the Rényi reflected entropies \(S_{R,\rho}^{(\alpha)}\) for \(0<\alpha<2\) is a correlation measure. The counterexamples are diagonal in a product basis and therefore already classical. The theorem is stated for \(\rho_{ABC}\) on \(\mathbb C^3\otimes\mathbb C^3\otimes\mathbb C^2\), and it directly undermines any naïve claim that a reflected-entropy-based multipartite object is automatically a universal correlation measure [2302.10208].

This no-go result does not prove that every conceivable genuine reflected multi-entropy is impossible. What it does show is that any such proposal must add structure beyond ordinary reflected entropy intuition. The later prescription-based \(GS_R\) does exactly that by imposing vanishing on lower-partite-entangled pure states and by mixing reflected multi-entropies with ordinary multi-entropies. Even so, the construction remains incomplete: there is no closed general-\(q\) formula, no holographic dual for \(GS_R\), no general proof of nonnegativity or monotonicity, and no recoverability theorem analogous to the ordinary Markov-gap identity [2507.15262].

Several adjacent programs should also be distinguished from genuine reflected multi-entropy proper. Symmetry-resolved genuine multi-entropy studies fixed-charge multipartite entanglement in Haar-random and graph states, but it does not define reflected entropy, multipartite reflected entropy, or a canonical-purification-based genuine reflected quantity [2511.00905]. L-entropy,
\[
\ell_{A_iA_j}=\min\{2S(A_i),2S(A_j)\}-S_R(A_i:A_j),
\qquad
\ell_{A_1\cdots A_n}=\left(\prod_{i<j}\ell_{A_iA_j}\right)^{\frac{2}{n(n-1)}},
\]
is a reflected-entropy-inspired genuine multipartite entanglement monotone built from upper-bound deficits of pairwise reflected entropy; it is presented as a serious surrogate for a genuine reflected multi-entropy, but not as a multipartite reflected entropy derived from one global canonical purification [2602.00617]. These parallel developments confirm that the field currently contains a family of reflected, reflected-inspired, and genuine multipartite constructions rather than a single universally accepted definition.

In this sense, genuine reflected multi-entropy is best understood as an active research concept with one explicit subtraction-based realization, several holographic precursors, and a set of sharp conceptual constraints. Its defining ambition is stable: to combine the canonical-purification logic of reflected entropy with the irreducibility criterion of genuine multipartite entanglement. Its formal realization, however, remains plural, partially axiomatized, and still open to revision.

Source: https://www.emergentmind.com/topics/genuine-reflected-multi-entropy