---
title: 'Weak Core: A Multidisciplinary Analysis'
url: https://www.emergentmind.com/topics/weak-core
type: topic
---

# Weak Core: A Multidisciplinary Analysis

Weak core is a polysemous technical term whose meaning depends on disciplinary context. In multiplex network analysis, it denotes the weakly connected region inside an elite-like core that links distinct core communities. In cooperative and continuum game theory, it denotes a stability concept defined by the absence of certain coalitional blockings. In matrix analysis and proper \(*\)-rings, it appears in the weak core inverse and its \(m\)-dependent, weighted, and central generalizations. In astronomy and astrophysics, related expressions such as *core-weak mode*, *weak hot core*, and *weak cool-core* describe physically weakened or intermediate core states rather than a single abstract structure [1412.6913] [1903.09819] [2301.08818] [2301.12342] [2504.06802] [2606.29988].

## 1. Terminological scope

The expression *weak core* does not identify a universal invariant across fields. Instead, current usage separates into at least four families. The first is graph-theoretic and concerns sparse bridge regions inside multiplex network cores. The second is coalitional and concerns whether a feasible allocation or strategy can be blocked by a coalition that improves all of its members. The third is algebraic and concerns generalized inverses extending the core inverse beyond index-one square matrices or into weighted rectangular settings. The fourth is observational and adjectival, in which a core is said to be weak, weakened, or intermediate because its emission, cooling, or chemical richness is reduced relative to a stronger reference state [1412.6913] [2603.17862] [2403.14196] [2504.06802].

This dispersion of meaning is not merely terminological. In network science, the weak core is a subgraph extracted by path constraints. In game theory, it is a solution set defined by the nonexistence of blocking coalitions. In matrix theory, it is part of a hierarchy of generalized inverses. In astrophysics, it is an observational classification or mode label. This suggests that domain qualifiers are essential whenever the term is used outside a narrowly defined literature.

## 2. Weak core in multiplex network analysis

The most literal use of *Weak core* as a named object appears in the study of elites in social multiplex networks. A multiplex system is written as
\[
{\cal M}={\cal M}({\cal G}_1,\ldots,{\cal G}_\mu),
\]
with layers sharing the same node set \(V\) but having different edge sets \(E_\alpha\). To focus on relations present in all selected interaction types, the analysis begins with the intersection graph
\[
{\cal G}_{\cap}=\bigcap_{\alpha\leq \mu}{\cal G}_\alpha.
\]
The elite-relevant core is then not the standard \(K\)-core but the generalized \(K\)-core, \(\mathbf{G}_K\)-core, defined as the maximal induced subgraph in which each node either has degree at least \(K\) or connects two nodes whose degree is at least \(K\). Operationally, one recursively removes any node whose degree is lower than \(K\) and that has at most one nearest neighbor of degree at least \(K\). This construction preserves structurally important intermediaries that a standard \(K\)-core would discard [1412.6913].

Community structure inside \(\mathbf{G}_K({\cal G}_\cap)\) is isolated by applying the \(M\)-core, \(\mathbf{M}({\cal G})\), the maximal induced subgraph in which each link participates in at least \(M\) triangles. The connected components of
\[
\mathbf{MG}_K({\cal G}_\cap)
\]
are the *core communities*. Let these components be \(C_1,\ldots,C_m\). The weak core, \(\mathbf{W}_K({\cal G}_\cap)\), is then the subgraph of \(\mathbf{G}_K({\cal G}_\cap)\) formed by all nodes and links lying on a path that starts in some \(C_i\), ends in some \(C_j\) with \(i\neq j\), and whose intermediate nodes belong to
\[
\mathbf{G}_K({\cal G}_\cap)\setminus \mathbf{MG}_K({\cal G}_\cap).
\]
Its function is therefore explicitly connective: it is the part of the core that glues clustered core communities without itself being one of those communities. The paper also defines a minimal weak core \(\tilde{\mathbf{W}}_K\), obtained from minimal inter-community paths.

A further ingredient is the choice of the most relevant \(K\). This is based on the normalized Shannon entropy of the community-size distribution,
\[
p(C_j)=\frac{|C_j|}{\sum_{i\leq m}|C_i|}, \qquad
H(\mathbf{MG}_K)=-\sum_{i\leq m} p(C_i)\log p(C_i), \qquad
h(\mathbf{MG}_K)=\frac{H(\mathbf{MG}_K)}{\log m},
\]
with
\[
K^\star=\arg\max_K h(\mathbf{MG}_K).
\]
The intent is to select the \(K\) at which the core communities are most balanced, so that the bridging role of the weak core is most informative.

Empirically, the method was tested on the multiplex social network of the MMORPG *Pardus*, using friendship, communication, and trade layers over two 60-day windows. With \(M=2\), the generalized core exhibited three highly clustered core communities in both periods. These communities contained about \(68\%\) of the nodes of \(\mathbf{G}_K\) in the first period and \(61\%\) in the second, with \(K=14\) and \(K=13\) selected by the entropy criterion. The weak core had very low connectivity, clustering coefficient around \(c\sim 0.07\), and yet formed a single connected component. The minimal weak core was almost identical to the weak core itself, \(\tilde{\mathbf{W}}_K \approx \mathbf{W}_K\). Members of the weak core also had the best social performance indicators—experience, activity, age, and wealth—among the compared subgraphs, supporting the interpretation of the weak core as a sparse but socially central brokerage layer [1412.6913].

## 3. Weak core in cooperative and continuum games

In game theory, the weak core is a stability notion rather than a subgraph. In one-sided matching markets with endowments and no money, an allocation \(x\) is weak-core stable if no coalition \(C\) can *strongly block* it, meaning that there exists a feasible plan \(y\) such that
\[
\sum_{i \in C} (y_i - \omega_i) \le 0
\quad\text{and}\quad
y_i \succ_i x_i \ \forall i \in C.
\]
The essential feature is that every member of the blocking coalition must be strictly better off. This makes the weak core larger than the strong core. The 2026 analysis of multidimensional prices uses this notion as a baseline and shows that the rejective core is strictly stronger: every rejective-core allocation lies in the weak core, but not conversely. In the same framework, lexicographic dividend equilibria satisfy
\[
LDE_+(u,\omega)=\bigcap_{N=1}^\infty RC(u,\omega,N),
\]
and therefore lie inside the rejective core and hence also inside the weak core [2603.17862].

A different weak-core notion appears in normal form games with a continuum of players and without side payments. For payoffs depending on the full strategy profile, a coalition \(E\) blocks a strategy \(h\) if there exist \(\varepsilon>0\) and a coalition strategy \(f_E\) such that, for every complementary strategy \(f_{\complement E}\),
\[
U(t,f(t),f_E/f_{\complement E})>U(t,h(t),h)+\varepsilon
\quad \text{a.e. on } E.
\]
The weak core is the set of strategies not blocked in this sense. Under measurability and integrable boundedness of \(U\), together with concavity and equi-upper-semicontinuity in strategies, the weak core is nonempty, and an element of it is obtained as a limit of weak-core elements in appropriate finite approximating games. The same paper also shows that the weak core can be strictly larger than Aumann’s \(\alpha\)-core, and that in distribution-dependent continuum games, regularity conditions sufficient for pure-strategy Nash equilibria are irrelevant for weak-core non-vacuity [1903.09819].

Across these game-theoretic settings, *weak* has a precise logical role: blocking is weakened relative to stronger core notions, either by requiring strict improvement for all coalition members or by introducing an \(\varepsilon\)-margin approximation to the core.

## 4. Weak core inverse and matrix-theoretic generalizations

In linear algebra, weak-core terminology is attached to generalized inverses. The 2023 paper on the \(m\)-weak core inverse defines a new inverse for square matrices of arbitrary index by combining the \(m\)-weak group inverse with the orthogonal projector onto \(\mathcal R(A^m)\). One of its central formulas is
\[
A^{\circ_m}=(A^\oplus)^{m+1}A^m P_{A^m}.
\]
This object recovers the WC inverse when \(m=1\), the core-EP inverse when \(m\ge k=\operatorname{Ind}(A)\), and the classical core inverse when \(\operatorname{Ind}(A)\le 1\). It is also characterized as an outer inverse with
\[
\mathcal R(A^{\circ_m})=\mathcal R(A^k),
\qquad
\operatorname{rk}(A^{\circ_m})=\operatorname{rk}(A^k),
\]
and by systems such as
\[
XAX=X,\qquad AX=(A^\oplus)^m A^m P_{A^m},\qquad XA=(A^\oplus)^{m+1} A^m P_{A^m}A
\]
[2301.08818].

A further extension is the \(m\)-weak group MP inverse, built from the \(m\)-core-nilpotent decomposition. For \(A=A_1+A_2\), with \(A_1=AA^{\mathbb W_m}A\), the new inverse \(A^{\mathbb W_{m,t}}\) is the unique solution of
\[
XAX=X,\qquad AX=A_1A^\dagger,\qquad XA=A^\dagger A_1.
\]
It admits the representations
\[
A^{\mathbb W_{m,t}}=A^\dagger A_1 A^\dagger
\qquad\text{and}\qquad
A^{\mathbb W_{m,t}}=A^{\mathbb W_m}AA^\dagger.
\]
This construction unifies the weak core inverse at \(m=1\), the DMP-inverse when \(m\ge k-1\), and the \(m\)-weak group inverse in the EP case [2411.00022].

The rectangular weighted extension is the \(W\)-weighted \(m\)-weak core inverse. For \(A\in\mathbb C^{p\times n}\) and nonzero \(W\in\mathbb C^{n\times p}\), it is defined by the projection condition
\[
A^{\#m,W}=A^{\#m,W}P_{(WA)^m}.
\]
It reduces to the rectangular extension of the WC inverse at \(m=1\) and to the \(W\)-weighted core-EP inverse when \(m\ge k\). The theory includes explicit formulas through weighted Drazin and weighted core-EP inverses, range/nullspace characterizations, a canonical form via simultaneous unitary block upper triangularization of \((A,W)\), and applications to matrix equations such as
\[
x=A^{\#m,W}b+\bigl(I_p-A^{\#m,W}WAW\bigr)y
\]
for an associated constrained system [2403.14196].

In proper \(*\)-rings, the weak core inverse is defined by the existence of \(y\) and \(k\) such that
\[
ya^{k+1}=a^k,\qquad ay^2=y,\qquad (a^k)^*ay=(a^k)^*.
\]
The paper proves uniqueness, shows that weak core invertibility implies Drazin invertibility with matching index, derives formulas through Drazin and \(\{1,3\}\)-inverses, and establishes the power rule
\[
(a^n)^{wc}=(a^{wc})^n.
\]
It also introduces the *central weak core inverse*, obtained by adding the centrality condition \(ax\in C(R)\), together with additive laws under orthogonality assumptions [2005.13782].

## 5. Astrophysical and astrochemical usages

In observational astrophysics, weak-core language typically denotes a physically weakened or intermediate core state. For the pulsar PSR B0329+54, sensitive 2250 MHz observations identified a new *core-weak mode* in which the central core component becomes very faint while leading, trailing, and bridge components evolve in a coordinated sequence. The pattern lasts for \(3\) to \(14\) rotation periods, with most events lasting \(3\) periods and the longest barely exceeding \(14\); \(1759\) core-weak patterns were recognized. The duration distribution is positively skewed and fitted by a log-normal form with
\[
\sigma = 0.49\pm0.04,\qquad \mu = 1.40\pm0.02,\qquad A = 2662\pm324,
\]
and no periodicity was found in the occurrence of these events. Averaged over eight timescales, the mean amplitude ratios are
\[
\frac{P_1'}{P_1}=1.65,\qquad
\frac{P_2'}{P_2}=0.14,\qquad
\frac{P_3'}{P_3}=0.93.
\]
The observed sequence is not a simple disappearance of the core but a structured phase shift and intensity redistribution in which the core appears to drift out of and later return to the normal radiation window [2301.12342].

In high-mass star formation, *weak hot core candidates* are compact sources whose hot-core character is revealed only after weighted spectral stacking of multiple unblended transitions of complex organic molecules. In the ALMA-ATOMS survey, this method identified \(40\) new weak candidates in addition to \(60\) previously known strong hot cores. Classification required compact CH\(_3\)OH emission and at least one additional complex organic molecule among six other species. For the weak candidates, the column densities of complex organic molecules are approximately one order of magnitude lower than in the strong sample, but the full sample exhibits tight correlations between compact CH\(_3\)OH emission and other species, suggesting a shared chemical environment [2504.06802].

In galaxy-cluster astrophysics, a *weak cool-core cluster* denotes a system with a central cool component but without the sharply peaked structure of a classic strong cool core. XRISM observations of A3571 measured a nearly uniform core velocity dispersion of roughly \(100\)–\(120~\mathrm{km\,s^{-1}}\) out to \(120\) kpc, with a \(68\%\) upper limit of \(68~\mathrm{km\,s^{-1}}\) in the northern gas-sloshing elongation. The inferred 3D turbulent Mach numbers are \(0.16\pm0.02\) and \(0.14\pm0.01\) in the two radial bins, with non-thermal pressure fractions \(1.3\pm0.3\%\) and \(1.0\pm0.2\%\). Despite these low values, the derived turbulent heating rate is sufficient to offset radiative cooling losses, and sloshing motions are suggested to contribute significantly to the heating budget [2606.29988]. Related cluster work also studies the propagation of a weak shock of Mach number \(\sim 1.2\) through a cool core, emphasizing conductive temperature precursors and magnetic-field-dependent anisotropy rather than a weak-core classification [2004.13252].

## 6. Comparative interpretation

Across these literatures, the word *core* continues to indicate a privileged central object: an elite subgraph, a stable allocation set, a distinguished generalized inverse, or a physically central emitting or cooling region. The word *weak* then modifies that object in different ways. In multiplex networks, it denotes a low-connectivity and low-clustering region that is nevertheless indispensable for cohesion. In game theory, it denotes a relaxation of stronger blocking or stability requirements. In matrix theory, it marks an extension of stronger core-type inverses through projectors, power conditions, or weighted constructions. In astrophysics, it denotes reduced intensity, reduced chemical richness, or an intermediate thermodynamic state [1412.6913] [2603.17862] [2301.08818] [2301.12342].

This suggests a recurrent semantic pattern rather than a shared formalism. The weak core is typically not the densest, strongest, or most restrictive part of a system; instead, it is the residual or relaxed structure that remains central to connectivity, stability, inversion, or diagnosis. The term therefore has high internal coherence within each field but low transferability across fields. In practice, its meaning is determined almost entirely by the surrounding theoretical machinery: generalized \(K\)-cores and \(M\)-cores in multiplex networks, coalition blocking relations in economic theory, Drazin/core-EP/WC hierarchies in algebra, or phase-resolved and spectroscopic diagnostics in astrophysics.

Source: https://www.emergentmind.com/topics/weak-core