Monodromy Representation of Graphs

• Monodromy representation of graphs is a framework that encodes graph data using permutation actions, coverings, and homotopy techniques.
• The constructions include double-coset models, zigzag permutations in embedded graphs, and certified homotopy graphs for tracking monodromy actions.
• Applications span Hurwitz theory, polyhedral products, and singularity monodromy, offering powerful tools for topological and algebraic investigations.

to=arxiv_search.search 大发时时彩是json abstract=true author="" category="" ids="" max_results=10 query="all:\"monodromy representation of graphs\" OR ti:\"Monodromy representation of graphs\" OR ti:\"On Z-monodromies in embedded graphs\" OR ti:\"Certifying Galois/monodromy Actions via Homotopy Graphs\"" sort_by="submittedDate" sort_order="descending" title=""

to=arxiv_search.search шықәса 天天中彩票公众号json
{"abstract":true,"author":"","category":"","ids":"","max_results":10,"query":"all:\"monodromy representation of graphs\" OR ti:\"Monodromy representation of graphs\" OR ti:\"On Z-monodromies in embedded graphs\" OR ti:\"Certifying Galois/monodromy Actions via Homotopy Graphs\"","sort_by":"submittedDate","sort_order":"descending","title":""}
to=arxiv_search.search 天天中彩票是不是json
{"query":"ti:\"Monodromy representation of graphs\"","max_results":5,"sort_by":"submittedDate","sort_order":"descending","abstract":true}

Monodromy representation of graphs denotes a family of constructions in which graph-theoretic data are encoded by permutation actions, linear representations, or graph-local return maps attached to coverings, embeddings, or moduli. In current usage, the phrase is not confined to a single formalism. It includes double-coset models in which every graph is realized as a monodromy graph, local face permutations arising from zigzags in embedded graphs, homotopy-graph discretizations of Galois and monodromy actions of polynomial systems, monodromy representations attached to graph products via polyhedral products, dual-graph formulas for degeneration monodromy, and graph encodings of mapping classes and singularity monodromies [2509.17910] [2308.14123] [2603.17288] [1604.05504].

1. Double-coset realization of arbitrary graphs

In Yuan–Wang’s formulation, a graph is represented by group-theoretic data ((G;U,\rho,\tau)) with (G=\langle \rho,\tau\rangle), (\tau^2=1), and (U) a core-free subgroup. The associated monodromy graph (\Gamma=(G;U,\rho,\tau)) has vertex set
[
V={Ug\langle\rho\rangle\mid g\in G},
]
that is, double cosets of (U) and (H:=\langle \rho\rangle). Its edge set is
[
E=\big{{Uh\langle\rho\rangle,\ Uh\tau\langle\rho\rangle}\mid h\in S\big},
]
for a right transversal (S) of (U), and the multiplicity of the edge between (Uh\langle\rho\rangle) and (Uh\tau\langle\rho\rangle) is defined by
[
m\big(Uh\langle\rho\rangle,\ Uh\tau\langle\rho\rangle\big)=|D|.
]
The construction is intrinsic: if (T) is any set of right coset representatives for (U), then
[
E(\Gamma)={{UtH, Ut\tau H}\mid t\in T}={{UgH, Ug\tau H}\mid g\in G}.
]
The valency of a vertex satisfies
[
\mathrm{val}(UhH)=\frac{|\langle\rho\rangle|}{|\langle\rho\rangle\cap U^h|}.
]

This framework extends the classical coset-graph representation of vertex-transitive graphs. The central theorem states that every graph is isomorphic to a monodromy graph. For connected simple graphs with more than one vertex, the construction starts from the arc set (D), defines (\tau(u,v)=(v,u)), chooses (\rho) so that each (\langle\rho\rangle)-orbit is the set of arcs incident with a fixed vertex, sets (G=\langle\rho,\tau\rangle), and obtains
[
\Gamma=(G;G_\alpha,\rho,\tau)
]
isomorphic to the original graph. The same formalism also covers loops, multiple edges, and free edges [2509.17910].

The double-coset model is not only representational. If (U=1), then (\Gamma) is arc-transitive. Hence every graph (\Gamma=\mathrm{Mon}(G;U,\rho,\tau)) canonically gives rise to an arc-transitive graph
[
\Sigma=\mathrm{Mon}(G;1,\rho,\tau),
]
and to an orientable regular map (M(G;1,\rho,\tau)). Yuan–Wang summarize this by a “fundamental triad” in algebraic graph theory: where there is a graph, there is a group, an arc-transitive graph, and an orientable regular map. They also reinterpret algebraic maps as (M(G;U,\rho,\tau)), and prove an enumeration theorem: if (m) is the number of conjugacy classes of core-free subgroups of (G), and (G) admits (n) pairwise non-isomorphic regular maps, then the number of non-isomorphic maps admitted by (G) is (mn) [2509.17910].

A common misconception is that this double-coset formalism exhausts the topic. It does not. Other literatures use “monodromy representation of graphs” in substantially different local, topological, and computational senses.

2. Zigzag monodromy in embedded graphs

For a connected finite graph (T) embedded as a map in a connected closed surface (S), with the standing assumption
[
(SS)\quad T\text{ and the dual map } \mathcal I^* \text{ both are embeddings of simple graphs},
]
zigzags determine a local monodromy permutation attached to each face (F). If (F) is (k)-gonal, its oriented boundary edges form
[
\mathcal Q(F)={e_1,\dots,e_k,-e_k,\dots,-e_1},
]
with boundary rotation
[
D_F=(e_1,e_2,\dots,e_k)(-e_k,-e_{k-1},\dots,-e_1).
]
Given (e\in\mathcal Q(F)), let (e_0) satisfy (D_F(e_0)=e). There is a unique zigzag containing the consecutive pair ((e_0,e)), and (M_F(e)) is defined to be the first oriented boundary edge of (F) appearing again in that zigzag after the segment (e_0,e). This yields the (z)-monodromy
[
M_F:\mathcal Q(F)\to\mathcal Q(F).
]

Under the identification (\mathcal Q(F)\cong [k]^{\pm}), the corresponding permutation (\sigma) satisfies two structural constraints:
[
(M1)\quad \text{If } \sigma(i)=j,\text{ then } \sigma(-j)=-i,
]
[
(M2)\quad \sigma(i)\neq -i.
]
Lemma 1 gives the graph-theoretic content of these identities: reversibility under zigzag reversal, bijectivity, and the prohibition (M_F(e)\neq -e). The main theorem is a complete local classification: every permutation (\sigma) of ([k]^{\pm}) satisfying (M1) and (M2) is realized as the (z)-monodromy of a (k)-gonal face in a connected finite graph embedded in any connected closed surface, orientable or not, subject to (SS). There are no additional constraints on cycle types or parity. Section 4 constructs the plane case; Section 5 extends it to arbitrary surfaces by connected sums [2308.14123].

This is a genuinely local monodromy representation. It does not act on flags or on the full incidence structure, as in classical map monodromy groups. Instead, it records how zigzags recur to a given face. When (T) and (T^*) are simple, zigzags correspond to central circuits in the medial graph (M(T)), so (M_F) can equally be regarded as monodromy induced by central circuits of the medial graph acting on oriented boundary edges of (F). The classification shows that, under simplicity of both primal and dual, local zigzag monodromy is constrained only by reversal symmetry and the exclusion of immediate reversal [2308.14123].

3. Homotopy graphs and certified monodromy actions

For a parametrized polynomial system
[
F(x;z)=0
]
with generic fiber of size (d), monodromy is the representation
[
\pi_1(B,b_0)\longrightarrow \operatorname{Sym}(S_{b_0})\cong S_d,
]
where (B) is parameter space minus the branch locus and (S_{b_0}=\pi^{-1}(b_0)). The paper “Certifying Galois/monodromy Actions via Homotopy Graphs” discretizes this representation by a connected multigraph (\mathcal G=(V,E)) embedded in parameter space. Vertices are generic parameter points; each edge (e=(u,v)) is equipped with a path (p_e:[0,1]\to\mathbb C^m) and parameter homotopy
[
H_e(x;t):=F(x;p_e(t)).
]
Tracking solutions along (H_e) defines an edge correspondence
[
\gamma_e:\pi^{-1}(u)\longrightarrow \pi^{-1}(v).
]

Closed walks in (\mathcal G) based at (z_0) produce permutations by composition:
[
\sigma=\gamma_{e_r}\circ\cdots\circ\gamma_{e_1}\in S_k.
]
The resulting graph-based monodromy group is
[

\Mon_{\mathcal G,z_0}

\left\langle
\gamma_{e_r}\circ\cdots\circ\gamma_{e_1}
\;\middle|\;
(e_1,\dots,e_r)\in\mathcal P
\right\rangle
\subset S_k.
]
If (\mathcal G) is saturated and (k=d), Theorem 4.1 states that this coincides with the monodromy of the cover restricted to the region explored by the graph. A spanning tree (\mathcal T) yields explicit generators from off-tree edges:
[

\sigma_{\mathcal T,e}

\phi_{\mathcal T,v}^{-1}\circ\gamma_e\circ\phi_{\mathcal T,u},
]
and these generate (\Mon_{\mathcal G,z_0}) [2603.17288].

The distinctive feature of this construction is certification. The Krawczyk operator
[
K(F,x,r,A) := -AF(x) + (Id - A\square JF(x+rB))\, rB
]
and the theorem (K(F,x,r,A)\subset r\rho B) with (\rho<1) guarantee a unique root in the box (x+rB). CertifiedTrack uses interval arithmetic, predictors, refinement, and KrawczykTest to rule out path jumping. The correctness theorem then states that the computed generators are induced by actual analytic continuation along loops in parameter space. This turns monodromy into a rigorously computable graph action, and the framework is applied to generic univariate families, Elkies’ degree-23 Belyi polynomial with monodromy group (M_{23}), symmetric cubic surfaces, nearest-point problems, P3P, and the 5-point relative pose problem [2603.17288].

4. Graph products, polyhedral products, and free-group monodromy

A different usage arises from graph products of finite groups. Given a simplicial graph (\Gamma) with vertex groups (G_i), the graph product is
[

G_\Gamma

\left( G_1 * \cdots * G_n \right)
/ \langle!\langle [g_i,g_j] : {i,j}\text{ edge of }\Gamma\rangle!\rangle.
]
If (K) is a simplicial complex with 1-skeleton (K_1=\Gamma), the polyhedral-product fibration
[
(EG,G)^K \longrightarrow (BG,1)^K \longrightarrow \prod_{i=1}^n BG_i
]
has
[
\pi_1((BG,1)^K)\cong \prod_{K_1}G_i.
]
When (K) is flag and (K_1) is chordal, the fiber has free fundamental group (F_{p_K}), giving a short exact sequence
[
1\longrightarrow F_{p_K}\longrightarrow \prod_{K_1}G_i \longrightarrow \prod_{i=1}^n G_i \longrightarrow 1.
]
The conjugation action on the free kernel defines the monodromy representation
[
\Theta_K:\prod_{K_1}G_i\to \operatorname{Aut}(F_{p_K}),
]
and, after passing to the quotient,
[
P_K:G_1\times\cdots\times G_n\to \operatorname{Out}(F_{p_K}).
]

Theorem 1.1 states that, for finite discrete groups (G_i) and chordal (K_1), both (\Theta_K) and (P_K) are faithful. Theorem 1.2 further shows that if the (G_i) are finite abelian groups, then the induced linear representation lands faithfully in
[
\mathrm{SL}(p_K,\mathbb Z),
]
and for non-abelian groups in
[
\mathrm{GL}(p_K,\mathbb Z).
]
The paper gives explicit examples for right-angled Coxeter groups, including
[
1 \to F_5 \to \mathbb Z_2 * \mathbb Z_2 * \mathbb Z_2 \to \mathbb Z_2\times \mathbb Z_2\times \mathbb Z_2 \to 1,
]
where the generators act by explicit automorphisms of (F_5) and by explicit matrices in (\mathrm{SL}(5,\mathbb Z)) after abelianization [1604.05504].

Here the graph does not parametrize a moduli space. Instead, the graph controls the commutation relations in the base group and, through the polyhedral-product fibration, determines a monodromy action on a free group and its homology.

5. Dual graphs and degeneration monodromy

In log geometry, monodromy of degenerations becomes essentially combinatorial. For a log smooth degeneration over the standard log disc, the Kato–Nakayama Betti realization recovers the topology of the germ of the family from the log special fiber alone. In the case of curves, the relevant data are essentially equivalent to those encoded by the dual graph of a semistable degeneration, including the monodromy pairing and the Picard–Lefschetz formula [1802.02234].

If (X) is a nodal central fiber, its dual graph (\Gamma(X)) has vertices given by irreducible components and edges given by nodes, with weights (\nu(y)) attached to nodes. The nearby cycle sheaves satisfy
[
\Psi^q_{X/S}\cong \bigwedge^q M_{X/S}^{\mathrm{gp}}(-q),
]
so the log quotient (M_{X/S}^{\mathrm{gp}}) controls monodromy. In the curve case, Proposition 7.13 identifies the differential
[
d_2^{0,1}:H^0(X_{\log},\Psi^1_{X/S})\to H^2(X_{\log},\mathbf Q(1))
]
with the graph boundary map
[
d_1:C_1(\Gamma(X))\to C_0(\Gamma(X)).
]
Consequently,
[
E_\infty^{0,1}\cong H_1(\Gamma(X),\mathbf Z).
]

The graph pairing recovers the monodromy pairing, and the Picard–Lefschetz formula becomes a weighted graph formula:
[
\rho_\gamma(x)=x-\sum_y \nu(y)\,\langle x,v_y\rangle\, v_y.
]
The paper also establishes Kummer étale analogues, so the same combinatorial monoid data govern both Betti and étale monodromy. In this regime, “monodromy representation of graphs” is literal: the dual graph carries the chain complex, pairings, and weighted incidence data from which the monodromy operator on (H^1) is reconstructed [1802.02234].

6. Monodromy graphs in Hurwitz theory

Hurwitz numbers count branched covers of (\mathbb P^1) with prescribed ramification, equivalently factorisations in (S_d). For type ((g,\mu,\nu)), one has
[
b=2g-2+\ell(\mu)+\ell(\nu),
]
and factorisations
[
(\sigma_1,\tau_1,\dots,\tau_b,\sigma_2)
]
with (\sigma_1,\sigma_2) of cycle types (\mu,\nu), (\tau_i) transpositions, product constraint
[
\sigma_2\cdot \tau_b\cdot \dots \cdot \tau_1\cdot \sigma_1 = \mathrm{id},
]
and transitivity condition. Construction 2.10 converts such data into a weighted graph (\Gamma\to [0,b+1]): in-ends over (0), out-ends over (b+1), one inner vertex over each integer (1,\dots,b), and cut/join behavior determined by the transpositions. A monodromy graph is connected, has first Betti number (g), satisfies balancing at each inner vertex, and its edge weights are the corresponding cycle lengths [1703.05590].

This graph is a tropical cover. The tropical simple double Hurwitz number is
[

\operatorname{Th}^{(2)}_{g;\mu,\nu}

\sum_{\Gamma}
\frac{1}{|\operatorname{Aut}(\Gamma)|}\prod_e w(e),
]
and Theorem 2.14 states
[
h^{(2)}{g;\mu,\nu}=\operatorname{Th}^{(2)}{g;\mu,\nu}.
]
Triply interpolated Hurwitz numbers refine the construction by coloring edges as normal, bold, or dashed and by attaching counters. These decorations encode weakly monotone, strictly monotone, and unconstrained portions of the transposition sequence. Reduced monodromy graphs then support Ehrhart-theoretic analysis of piecewise polynomiality and wall-crossing in genus (0), and more generally weighted lattice-point counts in higher genus [1703.05590].

In this setting, the graph is a compressed record of the full monodromy representation of the branched cover. Loops around branch points are replaced by cut/join vertices, and the symmetric-group action is replaced by a weighted tropical graph over an interval.

7. Tête-à-tête graphs, mixed twists, and singularity monodromy

A ribbon graph (\Gamma) embedded as a spine of an oriented surface with boundary (\Sigma) supports safe walks: unit-speed walks that, at each vertex, continue along the next edge in the cyclic order. For a metric ribbon graph, the (\ell)-tête-à-tête property requires that for any interior point (p), the two (\ell)-safe walks (\gamma_p) and (\omega_p) end at the same point, and that safe walks starting on the distinguished boundary subgraph (A) return to (A). This defines a homeomorphism
[
\sigma_\Gamma(p)=\gamma_p(\ell)
]
of the graph, and a tête-à-tête twist on the thickening surface. In the signed version, the sign function (\iota) assigns (+), (-), or (0) to boundary components not in (A), and the induced fractional Dehn twist coefficient is
[

\operatorname{rot}{\partial^1}(\phi{(\Sigma,\Gamma,A,\iota)},C_i)

\iota(i)\,\frac{\pi}{\mathrm{length}(\widetilde{\Gamma}_i)}.
]
The characterization theorem states that orientation-preserving mapping classes fixing the boundary pointwise and boundary-free isotopic to periodic homeomorphisms are exactly the signed tête-à-tête twists; the unsigned case corresponds to strictly positive fractional Dehn twist coefficients [1706.05580].

Mixed tête-à-tête graphs extend this to pseudo-periodic homeomorphisms. A filtered graph
[
(\Gamma,A)= (\Gamma^0,A^0)\supset (\Gamma^1,A^1)\supset\cdots\supset (\Gamma^d,A^d)
]
together with locally constant lengths (\delta_i) defines mixed safe walks by concatenating safe walks level by level. The mixed tête-à-tête property ensures that the resulting map
[
\phi_{(\Sigma,\Gamma^\bullet,\delta_\bullet)}
]
is well defined and pseudo-periodic. The realization theorem states that if the adjacency graph of periodic pieces is a tree, all screw numbers are nonpositive, and at least one fixed boundary component has positive fractional Dehn twist coefficient, then the mapping class is represented by a mixed tête-à-tête twist. For irreducible plane curve singularities, the monodromy of the Milnor fibration satisfies these hypotheses, hence is represented by a mixed tête-à-tête graph. For the singularities (x^p-y^q=0), the graph is the complete bipartite graph (K_{p,q}), and the induced tête-à-tête twist is the monodromy fixing the boundary of the Milnor fiber [1706.05580] [1712.05988].

A stronger statement is proved in the surface-singularity setting: mixed tête-à-tête twists are precisely the monodromies associated with reduced function germs on isolated surface singularities. The converse realization is equally explicit: given a mixed tête-à-tête graph (\Gamma), the cone over the open book associated with (\Gamma) admits a complex structure, and there exists a reduced holomorphic function germ
[
f:C(\Gamma)\to\mathbb C
]
whose Milnor fiber and monodromy realize the mixed tête-à-tête twist. In a sequel paper, the fourth author and B. Sigurdsson have extended this to the reducible case [1712.05988].

These constructions make the phrase “monodromy representation of graphs” especially literal: the graph is not merely an auxiliary combinatorial device but a complete encoding of the mapping class or singularity monodromy, including periodic pieces, Dehn twists, screw numbers, and boundary rotation data.