Inductive construction of path homology chains and the structure of $\Omega_3(G;R)$

Matthew Burfitt; Tyrone Cutler

REVIEW 4 major objections 3 minor 15 references

A recursive extension procedure over face multigraphs generates all path-homology chains in characteristic 2 and, through dimension 3, in characteristic 0.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 03:23 UTC pith:Z3YX6UMO

load-bearing objection The face-multigraph/inductive-element machine is a genuine advance for path homology, and the main theorems are believable, but Lemma 4.4 needs a written replacement argument before I would take Theorem 1.1 as fully established. the 4 major comments →

arxiv 2607.13906 v1 pith:Z3YX6UMO submitted 2026-07-15 math.AT

Inductive construction of path homology chains and the structure of Ω₃(G;R)

Matthew Burfitt , Tyrone Cutler This is my paper

classification math.AT MSC 05C2055N35

keywords path homologydigraph topologyface multigraphsinductive elementstrapezohedron elementsuniversal coefficient theoremGLMY theorycharacteristic 2

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper introduces a recursive way to build path-homology chains of a digraph from chains two dimensions lower, using upper and lower extensions parametrized by labelled multigraphs called face multigraphs. It claims that the resulting inductive elements generate every Omega_n(G;R) when R has characteristic 2, and generate Omega_n(G;R) for n up to 3 when R has characteristic 0. For dimension 3, it yields explicit generators obtained as images of trapezohedron elements under digraph maps, answering an open question for arbitrary digraphs without the usual restrictions on multisquares. The construction also transfers bases between rings of the same characteristic and produces universal coefficient statements in low degrees. If correct, this is the first chain-level description of path homology that works for all digraphs and for a broad class of coefficient rings.

Core claim

The paper establishes that strongly connected inductive elements—formed inductively from the vertex basis by complete extensions over face multigraphs—generate Omega_n(G;R) when R has characteristic 2 for every n, and when R has characteristic 0 for n ≤ 3. In particular, Omega_n(G;Z2) has a basis of inductive elements for every n, and Omega_n(G;Z) is generated by inductive elements for n ≤ 3. For dimension 3, the paper gives a generating set consisting of images of 3-path elements under 3-path structure maps, equivalently images of trapezohedron elements under digraph maps; this holds for any digraph G and any ring of characteristic 0 or 2. When G has no multisquares, these inductive element

What carries the argument

The central object is the upper (or lower) extension [x]_v of an n-chain x, obtained by appending a vertex v to every path in x, together with a labelled multigraph called a face multigraph whose vertices are lower-dimensional chains and whose edges record cancellations among their boundary components. A complete extension over such a face multigraph guarantees that the resulting element lies in Omega_{n+1}. Restricting to connected inductive structures defines inductive elements, and the load-bearing step is the claim that any element possessing one inductive structure decomposes as a sum of strongly connected inductive elements.

Load-bearing premise

The whole generating result depends on the claim that any chain carrying one recursive construction can be split into pieces each having only connected recursive constructions; this splitting step is asserted in a single sentence and is what makes inductive elements generate Omega_n.

What would settle it

Construct a digraph with an integral 4-chain that has at least one inductive structure but whose every inductive structure is disconnected, or find an integral 4-chain not expressible as a sum of strongly connected inductive elements; either would falsify the decomposition lemma and collapse the generating-set claims for characteristic 0.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

Omega_*(G;Z2) admits a basis of inductive elements for every digraph G, and Omega_n(G;Z) is generated by inductive elements for n = 0, 1, 2, 3.
Omega_3(G;R) is generated by images of trapezohedron elements under digraph maps whenever R has characteristic 0 or 2, with no restriction on the digraph.
For digraphs without multisquares, the inductive elements form bases over every ring of zero or prime characteristic, extending previous field-only basis descriptions.
Universal coefficient exact sequences hold through degree 2 for characteristic-2 coefficients; in degree 3 a single obstruction group B_3(G;R) measures the failure of the expected sequence.
Change-of-coefficient maps give isomorphisms HP_*(G;R) ≅ HP_*(G;Z) ⊗ R for characteristic-0 rings, and the paper obtains new universal coefficient isomorphisms for digraphs without multisquares over odd characteristic rings.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

A natural next test is whether inductive elements generate Omega_n(G;Z) for all n ≥ 4; the low-dimensional successes suggest that any failure would be tied to torsion or to the decomposition lemma rather than to the extension construction itself.
The uniqueness of inductive structures for 3-path elements up to symmetries suggests that these constructions could support an efficient algorithm for computing persistent path homology in dimensions 2 and 3, where no canonical basis was previously available.
Because face multigraphs decouple the chain construction from any particular choice of basis, the same extension formalism could plausibly be adapted to other homology theories defined through path or magnitude chains, though the paper does not pursue that transfer.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

The face-multigraph/inductive-element machine is a genuine advance for path homology, and the main theorems are believable, but Lemma 4.4 needs a written replacement argument before I would take Theorem 1.1 as fully established.

read the letter

Worth reading. The paper’s main contribution is the face-multigraph and inductive-element construction, and it is not a repackaging of Grigor’yan or Fu–Ivanov: those earlier bases required no-multisquare or no-double-edge hypotheses and field coefficients, while this construction is free of digraph restrictions and works integrally in low dimensions and over characteristic-2 rings in all dimensions. The Ω3 generator description via 3-path sequences and the obstruction group for the universal coefficient theorem are solid byproducts. Lemma 3.1 and Proposition 4.3 are genuinely proven, and Theorem 5.2 is a real proof, not a sketch.

The soft spot is Lemma 4.4. The proof says to choose an inductive structure with all connected components strongly connected, but that is exactly what needs proving. The missing argument is the replacement step: if a component had a disconnected inductive structure, you should be able to splice that structure into the original one and get strictly more components, contradicting maximality. I think the argument goes through—v-completeness is component-wise, and the upper-extension map is injective for these heads—but it is not in the text, and Theorem 1.1 depends on it. A referee should require that paragraph.

Two smaller issues. Proposition 7.1 delegates a homology computation to “best verified by direct computation,” and Example 7.2 leans on a GitHub repository. Both are probably fine, but the computation should be pinned down or the code version stated. Also, the authors’ own preprint [2] has almost the same title and is not positioned relative to the present claims; since it is their own work, the novelty boundary needs one explicit sentence.

Overall the central argument holds up. The paper deserves a serious referee and likely minor-to-moderate revision, not rejection.

Referee Report

4 major / 3 minor

Summary. The paper introduces an inductive method for constructing path homology chains: from elements of Ω_n and Ω_{n−1}, upper/lower extensions over labelled face multigraphs produce elements of Ω_{n+1}. It defines inductive elements as strongly connected extensions and proves, under coefficient restrictions, that they generate Ω_n(G;R): for characteristic 2 in all dimensions, and for characteristic 0 in dimensions at most 3. It then introduces 3-path sequence digraphs and 3-path structure maps to give an explicit description of Ω_3(G;R) in terms of induced images of trapezohedron elements, answering a question of Grigor'yan for rings of characteristic 0 or 2. Several universal coefficient statements for path homology are derived as applications.

Significance. If the main claims hold, this is a substantial advance in GLMY theory: it provides the first general chain-level generating set for path homology of arbitrary digraphs, including those with multisquares, and it gives explicit generators for Ω_3 over a broad class of coefficient rings. The paper is largely self-contained, with direct proofs of Lemma 3.1 and Proposition 4.3, a detailed inductive argument in Theorem 5.2, and reproducible computational code for the worked examples. The explicit 3-path structure maps and the low-dimensional universal coefficient analysis are valuable contributions. The main risk is that the central generation theorem rests on Lemma 4.4, whose proof is a single unsubstantiated sentence; several supporting case analyses in Section 7 are also omitted. These issues are fixable in principle but are load-bearing for the paper's central claims.

major comments (4)

[§4.4, Lemma 4.4] The proof of Lemma 4.4 is a single sentence: 'Among all inductive structures, there is one with all connected components strongly upper connected.' This is not established and the lemma is load-bearing: Theorem 5.2, Corollaries 5.4 and 5.5, and hence Theorem 1.1 all need the decomposition of an element with an inductive structure into strongly connected inductive elements, not merely elements with a connected structure. A correct proof requires a replacement argument: if a connected component F_r has total element y_r and y_r admits a disconnected inductive structure F'_r, one must show that F'_r ∪ ⋃_{q≠r} F_q is again an inductive structure on the original x with strictly more connected components. This requires checking that F'_r has the same head h(x), that its vertex labels lie in E_{n−1}, that its ridge labels lie in E_{n−2}, and that h(x)-completeness is preserved under disjoint un
[§5.1, Theorem 5.2 proof] The proof of Theorem 5.2 contains consistent index errors that obscure the face-multigraph construction. In the paragraph following equation (5.2), x_i is an element of Ω_{n−1}(G;R), so the expression δ^h_{n,v}(x_i) is not defined; the intended map is δ^h_{n−1,v}. The same error appears in the definition of V_x and in the subsequent sums. Moreover, the displayed identity '0=∂^M_{n−1,n−1}δ^h_{n,v}(x)=∑_{v∈V_x} ...' sums over v inside an expression already fixed in v; as printed it is not checkable. This cancellation/pairing step is the core of the proof that the constructed multigraph is h(x)-complete, so the proof must be rewritten with consistent indices and a clear explanation of why the only possible cancellations are between paired basis elements (5.4). The final sentence of the theorem also claims a basis of strongly connected elements over Z_2, but the proof as written only constru
[§7, Proposition 7.1] Proposition 7.1 is foundational for Theorem 7.3 and hence for the description of Ω_3(G;R), but its proof is incomplete. For G = L^u_t and L^l_t the proof is simply 'left to the reader'; for G = L_t the homology computation H_2(L_t;R) = 0 is dismissed as 'best verified by direct computation'. In addition, the claim that L_t is the unique generator of Ω_3(L_t;R) is justified by 'no other upper inductive structures can be formed' together with Corollary 5.4, but no enumeration or argument is supplied, and Corollary 5.4 itself depends on the unproved Lemma 4.4. Since Proposition 7.1 is the concrete foundation for all later 3-path structure-map results, the proof should be completed for all three families, and the H_2 computation at least summarized or relegated to a detailed appendix.
[§7, Theorem 7.4] Theorem 7.4 is the bridge from arbitrary inductive structures on I ∈ Ω_3(G;R) to an explicit 3-path structure map φ_I. The proof contains several substantial omitted case analyses: for line face multigraphs with m ≥ 2, for the four possibilities when both x_1 and x_m are directed squares, and for the verification of the injectivity conditions in parts (4)–(6) of Definition 7.6. Phrases such as 'the details are straightforward and left to the reader' and 'the details in each case are straightforward' cover a large number of configurations. Given that Theorem 7.4 is used to prove uniqueness up to the Sym(H) action and to justify the trapezohedron-generator description in Corollary 7.10, the proof needs to be expanded or organized into a systematic case table with all nonzero checks recorded.

minor comments (3)

[§8.2, Example 8.2] The displayed definition of I^t_4 contains an unreadable passage of non-text symbols ('A+⋯+A ⌟⟨⟨⟪rl⟫l⟩⟩...'). As printed, the construction cannot be verified. Moreover, the assertion that 'all face multigraphs labelled by inductive elements over which I^t_4 upper extends are connected' is stated without proof. This example is not central to Theorems 1.1 or 1.4, but it should be repaired before publication.
[§4.2, Definition 4.5] In Definition 4.5(1), the text says 'Its vertex labels x_1, . . . , x_2 are elements of E_{n−1}' — the final index should be m, not 2.
[§1, Theorem 1.5(2)] The introduction states that for n ≥ 3 no sequence of the form (1.1) can be exact in general, relying on examples from [8] and the authors' companion paper [2]. Since [2] is a self-citation and the counterexample is used as a negative statement, it would help readers to include a precise reference or a brief statement of the counterexample.

Circularity Check

0 steps flagged

No significant circularity: the main generation theorem is derived from definitions; the only self-citation is peripheral and non-load-bearing.

full rationale

The central derivation chain is self-contained: Definition 4.1–4.5 introduce extensions, face multigraphs, completeness, and strong connectedness; Proposition 4.3 proves that complete extensions lie in Ω_{n+1}; Theorem 5.2 constructs inductive structures on basis elements using only Lemma 3.1, Lemma 3.8, and the bigrading decomposition; Corollaries 5.4 and 5.5 then obtain generation by induction. No fitted parameter or predetermined target is renamed as a prediction. The proof of Lemma 4.4 in Section 4.4 is under-justified—it asserts in one sentence the existence of an inductive structure whose connected components are strongly connected, omitting the maximal-component replacement argument—but this is a proof gap, not circularity: the lemma's conclusion is not among the hypotheses. The only self-citation is to the authors' own [2, Example 6.2], used to exhibit failure of universal coefficient sequences in some dimensions; this does not support the main generation theorem, and the characteristic-2 counterexample is also attributed to independent work [8, §5.4]. The paper also openly lists unresolved cases in its 'Open questions,' consistent with a non-circular, partial result.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 4 invented entities

The paper introduces no fitted numerical parameters and depends on standard algebra or established path/magnitude homology results. The new objects are formal definitions rather than empirical postulates, so no free parameters or invented empirical entities are at play.

axioms (4)

domain assumption Path chain modules Omega_n(G;R) are identified with diagonal magnitude homology ker(d^M_{n,n}) via Asao's isomorphism (Lemma 2.1).
This identification is the foundation for Lemma 3.1 and for the face-map characterization of Omega_n used throughout the paper.
domain assumption Connected elements with fixed endpoints span Omega_n and give the bigraded decomposition Omega_n = direct sum Omega^{v_t,v_h}_n (Section 3.3, following [12, Lemma 2.2]).
The bigrading is used in Definition 3.2, in Lemma 3.8, and to prove Proposition 2.2 and Theorem 5.2.
standard math Over R=Z or R=Z_2, submodules of free modules are free and bases respect the bigrading; standard PID facts apply.
Theorem 5.2 selects bases B_{n-1}, B_{n-2}, B_n and uses uniqueness of basis decompositions. This is standard module theory over PIDs.
standard math The classical universal coefficient theorem for chain complexes of free abelian groups applies to the path chain complex after passing to free resolutions.
Propositions 3.7, 6.5, and Theorem 7.11 apply [15, Theorem 7.55 and Corollary 7.56].

invented entities (4)

Face multigraphs and upper/lower extensions no independent evidence
purpose: Bookkeeping devices that encode the cancellations needed for an (n+1)-chain built from n- and (n-1)-chains to lie in the path chain module.
New formal definitions (Definitions 4.2–4.4); they are internal tools, not externally testable entities.
Inductive elements and inductive structures no independent evidence
purpose: Generating sets for Omega_n in characteristic 2 and for low dimensions in characteristic 0.
Defined recursively in Definition 5.1 from face multigraphs; no external handle.
3-path sequence digraphs L_t, L^u_t, L^l_t and 3-path structure maps no independent evidence
purpose: Model digraphs whose induced images generate Omega_3(G;R).
New families introduced in Definitions 7.1–7.6; internal mathematical devices.
Obstruction group B_3(G;R) no independent evidence
purpose: Measures the failure of the universal coefficient exact sequence for path homology in degree 3 over characteristic-2 rings.
Defined in Theorem 7.11 as coker(mu^B_3); internal algebraic invariant.

pith-pipeline@v1.3.0-alltime-deepseek · 3712 in / 3978 out tokens · 232114 ms · 2026-08-02T03:23:14.663693+00:00 · methodology

0 comments

read the original abstract

Path homology plays a central role in digraph topology and GLMY theory more generally. Unfortunately, the computation of the path homology of a digraph $G$ is a two-step process, and until now no complete description of even the underlying chain complex has appeared in the literature. In this paper we introduce an inductive method of constructing elements of the path homology chain modules $\Omega_n(G;R)$ from elements in the preceding two dimensions. This proceeds via the formation of what we call upper and lower extensions, that are parametrised by certain labelled multigraphs which we introduce and call face multigraphs. The inductive elements we construct generate $\Omega_*(G;R)$ when $R$ has characteristic $2$. With characteristic $0$ coefficients, the inductive elements at least generate $\Omega_i(G;R)$ for $i=0,1,2,3$. In low dimensions, the inductive elements coincide with the natural generators, and when the digraph contains no multisquares, the inductive elements coincide with the basis elements produced by Fu and Ivanov. Inductive elements provide a new concrete structure on the path chain complex that can be directly applied to understand path homology, under no restriction on the digraph $G$. We employ inductive elements to construct explicit generators of $\Omega_3(G;R)$ for a ring $R$ of characteristic $0$ or $2$, answering an open question posed by Grigor'yan. Several universal coefficient statements for path homology are obtained as a byproduct.

discussion (0)

Reference graph

Works this paper leans on

15 extracted references · 3 linked inside Pith

[1]

Asao,Magnitude homology and path homology, Bull

Y. Asao,Magnitude homology and path homology, Bull. Lond. Math. Soc.55(2023), no. 1, 375–398

2023
[2]

Bufitt and T

M. Bufitt and T. Cutler,Inductive construction of path homology chains, (2024), arXiv:2411.09501

Pith/arXiv arXiv 2024
[3]

Burfitt,Path homology with field coefficients, November 2024,https://github

M. Burfitt,Path homology with field coefficients, November 2024,https://github. com/MatthewBurfitt/Path-homology-with-field-coefficients.git

2024
[4]

Chowdhury and F

S. Chowdhury and F. M´ emoli,Persistent path homology of directed networks, SODA’ 18, SIAM, 2018, p. 1152–1169

2018
[5]

T. Dey, T. Li, and Y. Wang,An efficient algorithm for 1-dimensional (persistent) path homology, Discrete Comput. Geom.68(2022), 1102–1132

2022
[6]

S. Di, S. Ivanov, L. Mukoseev, and M. Zhang,On the path homology of cayley digraphs and covering digraphs, Journal of Algebra653(2024), 156–199

2024
[7]

Dimakis and F

A. Dimakis and F. M¨ uller-Hoissen,Discrete differential calculus: Graphs, topologies, and gauge theory, J. Math. Phys.35(1994), no. 12, 6703–6735

1994
[8]

Fu and S

X. Fu and S. Ivanov,Path homology of digraphs without multisquares and its comparison with homology of spaces, (2024), arXiv:2407.17001

Pith/arXiv arXiv 2024
[9]

Grigor’yan, R

A. Grigor’yan, R. Jimenez, Y. Muranov, and S.-T. Yau,On the path homology theory of digraphs and eilenberg–steenrod axioms, Homology Homotopy Appl.20(2018), no. 2

2018
[10]

Grigor’yan, Y

A. Grigor’yan, Y. Lin, Y. Muranov, and S.-T. Yau,Homologies of path complexes and digraphs, 2013, arXiv:1207.2834

Pith/arXiv arXiv 2013
[11]

,Homotopy theory for digraphs, Pure Appl. Math. Q.10(2014), no. 4, 919–674

2014
[12]

Grigor’yan,Advances in path homology theory of digraphs, ICCM10(2022), no

A. Grigor’yan,Advances in path homology theory of digraphs, ICCM10(2022), no. 2, 61–124

2022
[13]

Grigor’yan, Y

A. Grigor’yan, Y. Muranov, and S.-T. Yau,Homologies of digraphs and k¨ unneth formu- las, Comm. Anal. Geom.25(2017), 969–1018

2017
[14]

Hepworth and S

R. Hepworth and S. Willerton,Categorifying the magnitude of a graph, Homology Ho- motopy Appl.19(2017), no. 2, 31–60

2017
[15]

Rotman,An introduction to homological algebra, 2 ed., Springer, 2009

J. Rotman,An introduction to homological algebra, 2 ed., Springer, 2009. 63

2009

[1] [1]

Asao,Magnitude homology and path homology, Bull

Y. Asao,Magnitude homology and path homology, Bull. Lond. Math. Soc.55(2023), no. 1, 375–398

2023

[2] [2]

Bufitt and T

M. Bufitt and T. Cutler,Inductive construction of path homology chains, (2024), arXiv:2411.09501

Pith/arXiv arXiv 2024

[3] [3]

Burfitt,Path homology with field coefficients, November 2024,https://github

M. Burfitt,Path homology with field coefficients, November 2024,https://github. com/MatthewBurfitt/Path-homology-with-field-coefficients.git

2024

[4] [4]

Chowdhury and F

S. Chowdhury and F. M´ emoli,Persistent path homology of directed networks, SODA’ 18, SIAM, 2018, p. 1152–1169

2018

[5] [5]

T. Dey, T. Li, and Y. Wang,An efficient algorithm for 1-dimensional (persistent) path homology, Discrete Comput. Geom.68(2022), 1102–1132

2022

[6] [6]

S. Di, S. Ivanov, L. Mukoseev, and M. Zhang,On the path homology of cayley digraphs and covering digraphs, Journal of Algebra653(2024), 156–199

2024

[7] [7]

Dimakis and F

A. Dimakis and F. M¨ uller-Hoissen,Discrete differential calculus: Graphs, topologies, and gauge theory, J. Math. Phys.35(1994), no. 12, 6703–6735

1994

[8] [8]

Fu and S

X. Fu and S. Ivanov,Path homology of digraphs without multisquares and its comparison with homology of spaces, (2024), arXiv:2407.17001

Pith/arXiv arXiv 2024

[9] [9]

Grigor’yan, R

A. Grigor’yan, R. Jimenez, Y. Muranov, and S.-T. Yau,On the path homology theory of digraphs and eilenberg–steenrod axioms, Homology Homotopy Appl.20(2018), no. 2

2018

[10] [10]

Grigor’yan, Y

A. Grigor’yan, Y. Lin, Y. Muranov, and S.-T. Yau,Homologies of path complexes and digraphs, 2013, arXiv:1207.2834

Pith/arXiv arXiv 2013

[11] [11]

,Homotopy theory for digraphs, Pure Appl. Math. Q.10(2014), no. 4, 919–674

2014

[12] [12]

Grigor’yan,Advances in path homology theory of digraphs, ICCM10(2022), no

A. Grigor’yan,Advances in path homology theory of digraphs, ICCM10(2022), no. 2, 61–124

2022

[13] [13]

Grigor’yan, Y

A. Grigor’yan, Y. Muranov, and S.-T. Yau,Homologies of digraphs and k¨ unneth formu- las, Comm. Anal. Geom.25(2017), 969–1018

2017

[14] [14]

Hepworth and S

R. Hepworth and S. Willerton,Categorifying the magnitude of a graph, Homology Ho- motopy Appl.19(2017), no. 2, 31–60

2017

[15] [15]

Rotman,An introduction to homological algebra, 2 ed., Springer, 2009

J. Rotman,An introduction to homological algebra, 2 ed., Springer, 2009. 63

2009