Twofold universality of large-$N$ melonic random tensors

Carlos I. Perez-Sanchez

REVIEW 1 cited by

Twofold universality of large-$N$ melonic random tensors

T0 review · reviewed 2026-07-10 · glm-5.2

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Melonic tensors forget their rank at large N

desk verdict Genuine universality result with a real gap in the multi-trace maximality argument read the letter →

arxiv 2607.08677 v1 pith:JYTBE5DN submitted 2026-07-09 math.CO cond-mat.stat-mechhep-thmath-phmath.MPmath.PR

Carlos I. Perez-Sanchez This is my paper · ORCID

classification math.COcond-mat.stat-mechhep-thmath-phmath.MPmath.PR MSC 05C3082B2060B99

keywords melonicgraphstensormodelslargeNlimituniversalityWickcontractioncolouredintegrals1/Nexpansion

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proves that integrals of products of melonic tensor invariants—a class of graph-encoded contractions of tensor indices built by recursive dipole insertion—have a leading large-N behavior that depends neither on the tensor rank D (for D≥3) nor on the combinatorial details of the specific melonic graphs involved, but only on how many vertices each graph has. The author constructs a universal measure on rank-3 tensors that reproduces, at large N, the cumulants of any melonic tensor model of any rank D≥3, provided the coupling constants and vertex counts match. The mechanism rests on showing that melonic graphs admit a unique maximal Wick contraction, and that the number of ways to combine these contractions across multiple graphs into a connected structure is itself a combinatorial invariant depending only on vertex counts. A family of melonic polynomials is introduced that conjecturally gives the exact leading-order values of these integrals.

What carries the argument

The argument turns on three load-bearing pieces: (1) Lemma 3.1, which shows that any maximal Wick contraction must pair the two vertices of any dipole; (2) Proposition 3.6, which uses this to prove uniqueness of the maximal Wick contraction π_max for any connected melonic graph; and (3) Proposition 3.13, which characterizes the set of maximal connected Wick contractions of multiple melonic graphs as the set G_n defined by two conditions—that the contraction is obtained by swaps from the product of individual maximal contractions (condition i), and that a contracted 'thin graph' is a tree (condition ii). The face-count formula in Lemma 3.12, which gives the maximum number of faces as (D−1)(P−

What would settle it

Find a melonic graph B (or a collection B₁,...,Bₙ) and a Wick contraction Π that maximizes the number of faces but whose associated thin graph is not a tree, which would contradict Proposition 3.13 and break the chain leading to Theorem 4.2. Alternatively, find a case where the leading order of a melonic integral depends on D≥3 or on the combinatorial structure of the melonic graphs beyond vertex counts, which would directly contradict Theorem 4.2.

Watch

Extended reading notes

Core claim

The central result is that the rescaled connected integral of a product of melonic tensor invariants B₁(T)···Bₙ(T) of rank D, when multiplied by the appropriate power of N, converges as N→∞ to a limit that is identical for all D≥3 and depends on the graphs Bᵢ only through their vertex counts 2pᵢ. This is proved by showing that (a) each connected melonic graph has a unique Wick contraction maximizing the number of faces, (b) the number of maximal connected Wick contractions of a disjoint union of melonic graphs is a function only of the vertex counts (Proposition 3.13 and Corollary 3.14), and (c) a canonical rank-3 melonic graph u_p with 2p vertices can therefore replace any rank-D melonic Bᵢ

Load-bearing premise

The proof that every maximal Wick contraction of multiple melonic graphs satisfies the tree condition (condition ii in the definition of G_n) proceeds by contradiction: if the thin graph is not a tree, one undoes swaps until it becomes one, claiming each undo strictly increases the face count. The intermediate graphs produced by undoing individual swaps are assumed to remain valid Wick contractions with well-defined face structure, but this is stated rather than fully argued.

Editorial extensions

If this is right

Any two melonic tensor models with the same coupling structure and vertex counts are indistinguishable at large N, regardless of their rank D≥3—this means the large-N phase diagram of a melonic tensor model is a universal object.
The critical locus of the quartic melonic tensor model is given by Σgᵢ + 1/4 = 0 for any D≥3 and any number of quartic interactions M≤D, extending the previously known equal-coupling case.
The melonic polynomials Mel_n(p₁,...,pₙ) = p₁···pₙ × (p₁+...+pₙ−1)_{n−2} conjecturally give closed-form leading-order values for all melonic multi-trace integrals, reducing a combinatorial enumeration to a Pochhammer-symbol expression.
Finite-N methods become essential for distinguishing tensor ranks, since all rank information is lost in the large-N limit for melonic observables.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the universal measure is rank-independent, then numerical simulations or bootstrap methods at finite N could extract rank-dependent corrections whose structure might reveal a systematic 1/N expansion organized by D, providing a finite-N fingerprint of the tensor rank.
The tree condition (ii) in the definition of G_n suggests a connection to combinatorial species or exponential generating functions: the count #G_n depending only on vertex counts hints that melonic multi-trace integrals might be expressible as coefficients of a single generating function indexed by n, analogous to the Harer-Zagier formula mentioned in the introduction.
The replacement of arbitrary melonic graphs by canonical ones u_p depending only on vertex count p suggests a decoupling between the 'topology' of the melonic graph (irrelevant at large N) and its 'size' (the only relevant parameter), which could extend to non-melonic graphs if a suitable analogue of unique maximal Wick contraction exists.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The headline: this paper proves that large-N melonic tensor integrals are independent of both the tensor rank D (for D≥3) and the fine-grained combinatorics of the melonic invariants—only vertex counts matter. That's a new and clean observation, and the single-trace case is solidly proved. The multi-trace case has a gap in a load-bearing position, and I think the stress-test concern about it is correct, or at least not obviously dismissible. Let me explain both sides briefly. The good: the proof chain for single melonic graphs is tight. Lemma 3.1 (dipole vertices must be Wick-contracted in maximal pairings) → Proposition 3.6 (unique maximal contraction) → Lemma 3.7 (face count) → Corollary 3.8 (integral equals 1) is clean and convincing. The construction of the universal melonic measure (Definition 4.1) and the reduction of arbitrary-rank integrals to rank-3 is a useful framework. The paper is also honest that Theorem-Conjecture 4.4 is conjectural and correctly notes it's not used downstream. The gap: Proposition 3.13 claims M_n = G_n, characterizing maximal Wick contractions of multiple melonic graphs. The M_n ⊂ G_n direction argues by contradiction: if the 'thin graph' isn't a tree, undoing l cycle-edge swaps increases faces by l×D (Eq. 3.20), contradicting maximality. But Eq. 3.17, which gives the -D face change per swap, is derived only for swaps between *different connected components*. For a cycle-edge swap, the two vertex pairs are already in the same component (connected through other swaps along the cycle). For each color c, the face containing (x,y') and the face containing (x',y) may already coincide, in which case undoing the swap splits one face (+1) rather than merging two distinct ones. The net face gain per color is in {-1,0,+1}, not necessarily +1. If enough colors have coincident faces, the total gain could be zero or negative, and the contradiction fails. This is in the critical path: Prop 3.13 → Cor 3.14 → Thm 4.2 → Thm 5.1. The result might still be true—the extensive numerical checks reported in Figure 1 support it—but the proof as written doesn't close the loop. The fix likely requires showing that cycle-edge swaps always gain a strictly positive number of faces (even if not exactly D), or a different characterization of maximality for the multi-trace case. This deserves a serious referee who can either patch the argument or identify a counterexample. I'd recommend conditional acceptance pending resolution of this gap.

Referee Report

0 major / 0 minor

Summary. The paper proves a

Significance. The paper introduces a universal melonic measure and melonic polynomials, proving that large-N melonic tensor integrals are independent of both the tensor rank D≥3 and the fine-grained combinatorics of the melonic traces. The core proof chain (Lemmas 3.1–3.12, Proposition 3.13, Theorem 4.2) is structurally coherent. The uniqueness of maximal Wick contractions for single melonic graphs (Proposition 3.6) is convincingly argued by induction using dipole reduction. The application to quartic models (Corollary 5.2) leverages the Dartois-Eynard-Nguyen solution [NDE15] to derive a D-independent critical locus and two-point function, which is a concrete, falsifiable prediction. The computational verification of melonic polynomials via feyntensor [Pér26b] for thousands of integrals (Figure 1) is a notable strength.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the derivation chain is self-contained with no self-definitional or fitted-input circularity

full rationale

The paper's central result (Theorem 4.2) is derived from first principles: the Gaussian tensor measure (Eq. 1.7-1.8), Wick's theorem (Eq. 1.9), the amplitude definition (Eq. 1.10), and a sequence of combinatorial lemmas about face counts in melonic graphs (Lemmas 3.1, 3.4, 3.7, 3.12, Prop. 3.13, Cor. 3.14). The universal melonic measure (Def. 4.1) is constructed from canonical graphs u_p, not fitted to the target result. The couplings t_p in Eq. (5.2) are defined as explicit sums of the original couplings g_k, not as fitted parameters. The only external citation in the load-bearing chain is [NDE15] (Dartois-Eynard-Nguyen), used solely in Corollary 5.2 as an application to obtain the two-point function for a specific quartic model — this is an independent, externally verifiable result by different authors, not a self-citation. The Theorem-Conjecture 4.4 (melonic polynomials) is explicitly flagged as not used in the rest of the paper ('we will not use the statement 4.4 in this article'), so even if it were circular, it would not affect the main derivation. The proof of Theorem 4.2 proceeds by direct computation: both the D-coloured and rank-3 integrals are evaluated via their respective Wick contractions, shown to have the same cardinality by Corollary 3.14, and the N-scaling factors cancel to yield zero difference. No step reduces to its own inputs by construction. The skeptic's concern about Eq. (3.20) in the M_n ⊂ G_n direction of Prop. 3.13 is a correctness/completeness issue (whether the face-counting argument is fully rigorous for same-component swaps), not a circularity issue — the argument does not assume its conclusion, it attempts to derive it from face-counting, even if the derivation may have a gap. This is a matter for correctness assessment, not circularity analysis.

Assumptions & free parameters 1 free parameters · 4 assumptions · 2 invented entities

The paper introduces two new entities (universal measure, melonic polynomials) both with falsifiable predictions. The free parameters t_p are determined by the original couplings, not fitted. The main ad hoc axiom is Condition (ii) for G_n, which is structurally necessary for the proof but whose necessity is only argued, not independently proven.

free parameters (1)

t_p (universal measure couplings) = t_q = Σ_{k: #V(B_k)=q} g_k (Eq. 5.2)
Not free parameters in the fitting sense; they are determined by the original model couplings g_k via Eq. (5.2). No parameter is tuned to make the universality hold.

assumptions (4)

standard math Wick/Isserlis Theorem for Gaussian tensor integrals (Eq. 1.9)
Standard result used to decompose tensor integrals into sums over Wick contractions.
standard math Bijection between unitary invariants and regularly edge-D-coloured vertex-bipartite graphs (§1.1)
Established result from [GR12] used to encode tensor invariants as coloured graphs.
domain assumption Gurău's 1/N expansion exists for tensor models (§1.2)
The paper states it does not directly rely on Gurău's deep results but uses the amplitude formula (Eq. 1.10) which is part of that framework.
ad hoc to paper Condition (ii) in Definition of G_n: the thin graph associated to a maximal Wick contraction is a tree
This is introduced in §3.2 as a defining condition for G_n and then used in Prop. 3.13 to characterise maximal contractions. Its necessity is argued by contradiction but not independently proven.

invented entities (2)

Universal melonic measure independent evidence
purpose: A measure on (C^N)^⊗3 that replaces arbitrary melonic tensor measures at large N
The measure is explicitly constructed (Def. 4.1) and its equivalence to original measures is proven (Thm. 5.1). It makes falsifiable predictions: melonic polynomials (Eq. 4.6) give specific large-N values checkable by computation.
Melonic polynomials Mel_n(p₁,...,pₙ) independent evidence
purpose: Closed-form expressions for large-N melonic integrals
Defined by Eq. (4.6) as p₁···pₙ × (p₁+...+pₙ-1)_{n-2}. Verified computationally for many cases (Fig. 1) using feyntensor. Falsifiable for any untested case.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Twofold universality of large-$N$ melonic random tensors." pith.science (2026). https://pith.science/paper/JYTBE5DN

@misc{pith2026260708677,
  author       = {Pith},
  title        = {Pith review of: Twofold universality of large-$N$ melonic random tensors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JYTBE5DN}},
  note         = {Machine review of arXiv:2607.08677}
}

abstract

We construct a measure that exhibits two aspects of a new type of universality and dramatically simplifies the integration of tensors $T_{a_1,a_2,\ldots,a_D} \in \mathbb{C}$ ($a_1,\ldots,a_D=1,\ldots,N$) at large $N$. In contrast to matrix integration, in which matrix traces canonically yield the integrand, tensors need additional information (equivalent to a $D$-coloured graph $B$) to contract their indices and form a tensor trace $B(T)$. We show that, whenever each $B_1,\ldots, B_n$ can be obtained by a recursive construction known as melonicity, then the leading order in $N$ of the integral of $ {B_1}(T) {B_2}(T) \cdots {B_n}(T) $ is independent of the -- often intricate -- combinatorics of the traces $B_i$, but also, to our surprise, independent of $D$ as far as $D\geq 3$. Instead, at large $N$, these integrals are some functions (indexed by $n$) of the number of vertices $2p_i$ of $B_i$ which we call melonic polynomials. Melonic traces cumulants with respect to any ('interacting') measure \[ \exp\Big\{-N^{D-1} \sum_{i=1}^m g_i {B_i}(T)\Big\} \mathrm{d}\mu_0(T) \quad (g_1,\ldots,g_m \in \mathbb{R}, \mathrm{d}\mu_0(T) =\text{the tensor Gaussian}) \] with each $B_i$ melonic, can be computed with our universal measure that replaces each $B_i$ by a canonical trace depending only on $p_i$. We prove that any two melonic tensor models are indistinguishable at large-$N$, independently of the number of tensor indices (first universality aspect), and of the fine-grainedness of their interactions (second universality), being a sufficient condition that the couplings (the parameters $g_i$ above) agree and their respective traces are monomials with the same degree in $T$.

Figures

Figures reproduced from arXiv: 2607.08677 by the authors.

**Figure 1.** (Left). Each of these tables of width n is associated with integers p1 ě p2, . . . ě pn, where pi is the depth of the i-th column (from left to right; e.g. to the diagram (4.9), p1 “ 5, p2 “ 3; for the single white diagram, p1 “ 7, p2 “ 1). Green means ‘proved by explicit computation’, an empty or an absent diagram means ‘conjectured’ and pink means ‘proved by Corollary 3.8’. For thousands of integrals of 20 and … view at source ↗

**Figure 2.** Some melons for D “ 3: (a) 6-vertices melons and (b) 10-vertices melons. The diagram 4.9 requires having computed integral of their products (not all of them are independent) [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗

**Figure 3.** (a) Critical locus 4g1 ` 4g2 ` 4g3 ` 1 “ 0 of the measure (5.7) (M “ 3, D ě 3), as follows from the Universality Theorem 5.1, using the Dartois-Eynard-Nguyen solution. (b) Exact solution m2pg1, g2q of the largeN two-point function in the case M “ 2. Observe we plot two aspects with different parameters, and are not meant to be compared; instead, observe that the projection of the truncated straight line that shows … view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Logic tree of this article, with implications denoted by ordinary arrows. ‚ Universality is not expected in other theories based on tensors like [BGS13, OPVW15, Pér18], since these break unitary invariance, but it is worth exploring to which point it does. 6.3. Combina…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Redundant moments of non-melonic random tensor models: simplifying the tensor bootstrap
math-ph 2026-07 accept novelty 6.0 of 10

At large N, every melonic observable’s moment and its Schwinger–Dyson equation depend only on its degree, in arbitrary (including non-melonic) tensor models.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages · cited by 1 Pith paper

[1]

Three-dimensional simplicial quantum gravity and generalized matrix models

Jan Ambj rn, Bergfinnur Durhuus, and Th \'o rdur J \'o nsson. Three-dimensional simplicial quantum gravity and generalized matrix models. Mod. Phys. Lett. A , 6(12):1133--1146, 1991

work page 1991
[2]

Random tensor models in the large N limit: Uncolouring the coloured tensor models

Valentin Bonzom, R a zvan Gur a u, and Vincent Rivasseau. Random tensor models in the large N limit: Uncolouring the coloured tensor models . Phys. Rev. , D 85:084037, 2012

work page 2012
[3]

Counting Tensor Model Observables and Branched Covers of the 2-Sphere

Joseph Ben Geloun and Sanjaye Ramgoolam. Counting Tensor Model Observables and Branched Covers of the 2-Sphere . Ann. Inst. Henri Poincaré Comb. Phys. Interact. , 1:77--138, 2014

work page 2014
[4]

Critical behavior of coloured tensor models in the large N limit

Valentin Bonzom, R a zvan Gur a u, Aldo Riello, and Vincent Rivasseau. Critical behavior of coloured tensor models in the large N limit . Nucl. Phys. B , 853:174--195, 2011

work page 2011
[5]

3D Tensor Field Theory: Renormalization and One-loop -functions

Joseph Ben Geloun and Dine Ousmane Samary. 3D Tensor Field Theory: Renormalization and One-loop -functions . Ann. Henri Poincaré , 14:1599--1642, 2013

work page 2013
[6]

Counting gluings of octahedra

Valentin Bonzom and Luca Lionni. Counting gluings of octahedra. Electron. J. Comb. , pages P3--36, 2017

work page 2017
[7]

Some structural and enumerative aspects of discrete surfaces and PL-manifolds

Valentin Bonzom. Some structural and enumerative aspects of discrete surfaces and PL -manifolds. Habilitation thesis (HDR), arXiv :2212.12200 , 2022

work page Pith review arXiv 2022
[8]

O(N) Random Tensor Models

Sylvain Carrozza and Adrian Tanas a . O(N) Random Tensor Models . Lett. Math. Phys. , 106(11):1531--1559, 2016

work page 2016

Show all 28 references

[9]

The large N factorization does not hold for arbitrary multi-trace observables in random tensors

R a zvan Gur a u, Felix Joos, and Benjamin Sudakov. The large N factorization does not hold for arbitrary multi-trace observables in random tensors . Lett. Math. Phys. , 115(4):93, 2025

2025
[10]

R a zvan Gur a u and James P. Ryan. coloured Tensor Models - a review . SIGMA , 8:020, 2012

2012
[11]

Quantum Gravity and Random Tensors

R a zvan Gur a u and Vincent Rivasseau. Quantum Gravity and Random Tensors . arXiv:2401.13510 , 2024

2024 arXiv
[12]

Colored Group Field Theory

R a zvan Gur a u. Colored Group Field Theory . Commun. Math. Phys. , 304:69--93, 2011

2011
[13]

The 1/N expansion of coloured tensor models

R a zvan Gur a u. The 1/N expansion of coloured tensor models . Ann. Henri Poincaré , 12:829--847, 2011

2011
[14]

The complete 1/N expansion of coloured tensor models in arbitrary dimension

R a zvan Gur a u. The complete 1/N expansion of coloured tensor models in arbitrary dimension . Ann. Henri Poincaré , 13:399--423, 2012

2012
[15]

Universality for Random Tensors

R a zvan Gur a u. Universality for Random Tensors . Ann. Inst. H. Poincare Probab. Statist. , 50(4):1474--1525, 2014

2014
[16]

The Euler characteristic of the moduli space of curves

John Harer and Don Zagier. The Euler characteristic of the moduli space of curves. Invent. Math. , 85:457--485, 1986

1986
[17]

Multi-critical behaviour of 4-dimensional tensor models up to order 6

Luca Lionni and Johannes Th \"u rigen. Multi-critical behaviour of 4-dimensional tensor models up to order 6 . Nucl. Phys. B , 941: 600--635, 2019

2019
[18]

Finite- N bootstrap constraints in matrix and tensor models

Samuel Lalibert\'e and Reiko Toriumi. Finite- N bootstrap constraints in matrix and tensor models. arXiv:2603.17364 , 2026

2026 arXiv
[19]

An analysis of the intermediate field theory of T4 tensor model

Viet Anh Nguyen, St \'e phane Dartois, and Bertrand Eynard. An analysis of the intermediate field theory of T4 tensor model. J. High Energy Phys. , 2015(1):1--17, 2015

2015
[20]

P\'erez-S\'anchez , Fabien Vignes-Tourneret , and Raimar Wulkenhaar

Dine Ousmane Samary , Carlos I. P\'erez-S\'anchez , Fabien Vignes-Tourneret , and Raimar Wulkenhaar. Correlation functions of a just renormalizable tensorial group field theory: the melonic approximation . Class. Quant. Grav. , 32(17):175012, 2015

2015
[21]

P \'e r ez-S\'anchez

Carlos I. P \'e r ez-S\'anchez . Melonic polynomials and tensors. (In preparation)
[22]

P \'e r ez-S\'anchez

Carlos I. P \'e r ez-S\'anchez . The full Ward-Takahashi Identity for coloured tensor models . Commun. Math. Phys. , 358(2):589--632, 2018

2018
[23]

P \'e r ez-S\'anchez

Carlos I. P \'e r ez-S\'anchez . Bratteli networks and the Spectral Action on quivers . arXiv:2401.03705 , 2024

2024 arXiv
[24]

P \'e r ez-S\'anchez

Carlos I. P \'e r ez-S\'anchez . Critical curve of two-matrix models ABBA , A\ B,A\ B and ABAB , Part I: Monte Carlo . arXiv:2603.25715 , 2026

2026
[25]

P \'e r ez-S\'anchez

Carlos I. P \'e r ez-S\'anchez . feyntensor – a feynman integrator for tensors, 2026. (Software written in SageMath.) 10.5281/zenodo.19661197 https://doi.org/10.5281/zenodo.19661197

2026 doi
[26]

P \'e rez-S \'a nchez, and Brayden Smith

Nathan Pagliaroli, Carlos I. P \'e rez-S \'a nchez, and Brayden Smith. Bootstrapping Tensor Integrals . arXiv:2604.19714 2026

2026 arXiv
[27]

S ageMath, the S age M athematics S oftware S ystem ( V ersion 10.8) , 2026

The Sage Developers . S ageMath, the S age M athematics S oftware S ystem ( V ersion 10.8) , 2026. https://www.sagemath.org

2026
[28]

Tensor model for gravity and orientability of manifold

Naoki Sasakura. Tensor model for gravity and orientability of manifold. Mod. Phys. Lett. A , 6(28):2613--2623, 1991

1991

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.

[1] [1]

Three-dimensional simplicial quantum gravity and generalized matrix models

Jan Ambj rn, Bergfinnur Durhuus, and Th \'o rdur J \'o nsson. Three-dimensional simplicial quantum gravity and generalized matrix models. Mod. Phys. Lett. A , 6(12):1133--1146, 1991

work page 1991

[2] [2]

Random tensor models in the large N limit: Uncolouring the coloured tensor models

Valentin Bonzom, R a zvan Gur a u, and Vincent Rivasseau. Random tensor models in the large N limit: Uncolouring the coloured tensor models . Phys. Rev. , D 85:084037, 2012

work page 2012

[3] [3]

Counting Tensor Model Observables and Branched Covers of the 2-Sphere

Joseph Ben Geloun and Sanjaye Ramgoolam. Counting Tensor Model Observables and Branched Covers of the 2-Sphere . Ann. Inst. Henri Poincaré Comb. Phys. Interact. , 1:77--138, 2014

work page 2014

[4] [4]

Critical behavior of coloured tensor models in the large N limit

Valentin Bonzom, R a zvan Gur a u, Aldo Riello, and Vincent Rivasseau. Critical behavior of coloured tensor models in the large N limit . Nucl. Phys. B , 853:174--195, 2011

work page 2011

[5] [5]

3D Tensor Field Theory: Renormalization and One-loop -functions

Joseph Ben Geloun and Dine Ousmane Samary. 3D Tensor Field Theory: Renormalization and One-loop -functions . Ann. Henri Poincaré , 14:1599--1642, 2013

work page 2013

[6] [6]

Counting gluings of octahedra

Valentin Bonzom and Luca Lionni. Counting gluings of octahedra. Electron. J. Comb. , pages P3--36, 2017

work page 2017

[7] [7]

Some structural and enumerative aspects of discrete surfaces and PL-manifolds

Valentin Bonzom. Some structural and enumerative aspects of discrete surfaces and PL -manifolds. Habilitation thesis (HDR), arXiv :2212.12200 , 2022

work page Pith review arXiv 2022

[8] [8]

O(N) Random Tensor Models

Sylvain Carrozza and Adrian Tanas a . O(N) Random Tensor Models . Lett. Math. Phys. , 106(11):1531--1559, 2016

work page 2016