{"id":"a537dff7-1140-41ac-828e-9f3ffff5bb6e","arxiv_id":"2607.16334","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A hypergraph Tutte polynomial THG with deletion–contraction is introduced; it equals a degree-dependent random-cluster partition function and is incomparable with the Bernardi–Kálmán–Postnikov polymatroid Tutte polynomial TP.","lead":"This paper introduces a new Tutte polynomial for hypergraphs, encoding multi-way connection structure in two variables and satisfying a deletion–contraction rule that stays inside the hypergraph class. It connects that polynomial to statistical-mechanics models and shows it is incomparable with a known polymatroid Tutte polynomial, resolving an open question.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"TP equalities in Prop. 6.1 and the [26] recursion are the load-bearing external assumptions; if TP(H3)=TP(H4) is wrong or the recursion doesn't match BKP, Cor. 6.2 collapses.","rationale":"The reader's weakest-assumption identification is correct: the central new result is the negative answer to BKP, and it depends on external, unverified TP computations and on a recursive characterization from a cited preprint. I did not find an internal inconsistency in the main constructions: the deletion–contraction exponents in Theorem 2.1 balance, the universality constants in Theorem 3.1 match, the convolution algebra in Section 3 is consistent, and the Section 5 pair examples are explicitly verifiable. Thus the soft spot is not in the core theory but in the Section 6 comparison. Because the repo has no commit hash and the TP polynomials are not given, reproducibility is genuinely inadequate for a headline negative answer. This supports the reader's CONDITIONAL verdict rather than a full rejection, since a single independent recomputation could settle it.","tokens_in":30957,"tokens_out":27845,"duration_ms":233304,"concrete_test":"Independently recompute TP(PH3) and TP(PH4), and also TP(PH1) and TP(PH2), directly from the original Bernardi–Kálmán–Postnikov activity definition in [7], without using the [26] recursion or the linked repository; compare with the values asserted in Prop. 6.1. Separately verify that the [26] recursion (Definition 6.1) reproduces those values on these four polymatroids. If TP(H3)≠TP(H4) or TP(H1)=TP(H2), then Prop. 6.1 and Cor. 6.2 fail.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The negative answer to the BKP question rests on Proposition 6.1 and Corollary 6.2, which assert that TP(H3)=TP(H4) while THG(H3)≠THG(H4). These are the decisive steps showing TP does not determine THG and hence that the characteristic polynomial is not a specialization of TP. The full THG polynomials for H3/H4 are displayed and checkable, and the THG(H1)=THG(H2) claim is internally plausible, but the TP side is not substantiated in the text: Table 1 lists only some differing coefficients for TP(H1) vs TP(H2), and the equality TP(H3)=TP(H4) is asserted with no TP polynomial shown. The paper defers to a GitHub repository with no commit hash. Additionally, the working definition of TP is the recursive characterization from [26] (Definition 6.1), and Remark 6.1 relies on the unproved claim that this recursion agrees with the original Bernardi–Kálmán–Postnikov definition. If the repository computation of TP(H3)=TP(H4) is erroneous, or if [26] does not agree with [7] on these polymatroids, then the incomparability conclusion and the negative BKP answer do not follow. Since Corollary 6.2 is a headline result, this is the most load-bearing risk in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a hypergraph Tutte polynomial T_HG via a subset sum with edge-degree exponents, and proves it satisfies deletion-contraction (Theorem 2.1), multiplicativity, and planar duality. It also defines a k-polymatroid Tutte polynomial T_k as a constant-weight specialization of a polymatroid polynomial of Chávez-Lomelí et al., and proves deletion-contraction, duality, characteristic-polynomial specializations, a universality/recipe theorem, and convolution formulas. For (k+1)-uniform hypergraphs, T_HG specializes to T_k on the associated polymatroid, giving hypergraph-level universality and convolution theorems. The paper then shows T_HG is equivalent to a degree-dependent random cluster partition function and incomparable with two other hypergraph Potts-type partition functions. Finally, using the recursive definition of the Bernardi–Kálmán–Postnikov polynomial T_P from [26], it presents explicit 4-uniform hypergraphs showing T_HG and T_P are incomparable, and derives that the characteristic polynomial is not a specialization of T_P, answering negatively a question in [7].","tokens_in":31352,"tokens_out":22832,"duration_ms":206427,"significance":"The central construction is natural and the self-contained proof of Theorem 2.1 is clean. If the Section 6 computations and the identification of the recursive T_P with [7] are correct, the paper makes a solid contribution to hypergraph Tutte theory, with useful recipe theorems and statistical-mechanics interpretations. The negative answer to the BKP question is a notable result. The paper also explicitly displays the full T_HG polynomials for one of the two example pairs, which is a strength; however, the decisive T_P computations are not included in the text and are deferred to an unpinned GitHub repository.","major_comments":[{"comment":"The incomparability and the negative answer to the BKP question rest on four computational claims: T_HG(H1)=T_HG(H2), T_P(H1)≠T_P(H2), T_P(H3)=T_P(H4), and T_HG(H3)≠T_HG(H4). Only the last is fully documented (displayed polynomials and Table 2). The equality T_HG(H1)=T_HG(H2) is asserted without supporting polynomials, Table 1 lists only a few differing coefficients for T_P(H1),T_P(H2), and the equality T_P(H3)=T_P(H4) is asserted with no T_P polynomial shown. The text defers to a GitHub repository with no commit hash. Since Corollary 6.2 depends exactly on these equalities, please include the complete computed polynomials (or verifiable certificates) and a versioned, archived reference to the code.","section":"Section 6, Proposition 6.1 and Corollary 6.2"},{"comment":"The working definition of T_P is the recursive characterization from [26], and Remark 6.1 states without proof that it agrees with the Bernardi–Kálmán–Postnikov definition in [7]. The negative answer to the BKP question is only as strong as this identification. If the recursion in [26] does not agree with [7] on these polymatroids, Corollary 6.2 would not answer the BKP question. Please provide a proof of the agreement, or at least state the precise theorem from [26] and verify its hypotheses on the examples. Relying on an unreviewed arXiv preprint for a headline result without such a check is a load-bearing gap.","section":"Section 6.1, Definition 6.1 and Remark 6.1"}],"minor_comments":[{"comment":"The notation V\\e is ambiguous when e is a multiset. Please clarify that V\\e means removing the support of e, not removing d(e) many elements.","section":"Section 2, Definition 2.3"},{"comment":"Since Proposition 3.1 underpins Theorem 3.1 and the later hypergraph universality theorem, please include the short verification that T_k is exactly N(P,ω) with constant weights ω(e)=k, so that [14, Prop. 7.1] applies directly.","section":"Section 3.1, Proposition 3.1"},{"comment":"The associativity proof contains a confusing reindexing step, especially the transition from sums over B and C to the final expression. Please rewrite this part for clarity.","section":"Section 3.4, Proposition 3.8"},{"comment":"The formula divides by (w^k−1)^{1/k}; for w^k=1 the expression is singular. Please state that the identity is understood by continuity, or restrict to the region where the root is nonzero.","section":"Section 5, Corollary 5.3"},{"comment":"Reference [40] is a self-cited unpublished manuscript. Please mark it clearly as unpublished and, where a specific result is used, state the proposition being invoked.","section":"Section 5, Remark 5.1"},{"comment":"For both tables, please state explicitly that all coefficients not listed agree between the two polynomials. For T_P(H3) and T_P(H4), at least the full polynomials or a certificate of equality should be provided.","section":"Section 6, Tables 1 and 2"}],"recommendation":"major_revision","confidential_remarks":"The main risk is that the headline negative answer to the BKP question depends on unverifiable external computations and an unreviewed recursive characterization of T_P. The rest of the paper's core theory appears sound and original. If the authors supply complete, reproducible computations and verify the [26] recursion against [7] on the relevant class, I would be willing to support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a worthwhile paper. It delivers a new Tutte polynomial THG for hypergraphs that satisfies deletion–contraction within the hypergraph class, with a clean proof (Thm 2.1) and a nice raft of standard properties—multiplicativity, duality (via [18]), and for uniform hypergraphs, universality and convolution. The associated k-polymatroid polynomial T_k is honestly presented as a constant-weight specialization of Chávez-Lomelí et al.'s N, so the novelty there is the universality and convolution theorems plus the clean interface with (k+1)-uniform hypergraphs. The Potts/random cluster section is also solid: the equivalence of THG with the degree-dependent random cluster partition function is explicit and checkable, and the incomparability examples there are displayed with full polynomials.\n\nThe soft spot is exactly where the reader's report puts it: Section 6. The claim that THG and TP are incomparable in distinguishing power, and the negative answer to the BKP question, rest on Proposition 6.1, whose TP side is not substantiated in the text. For H1/H2 you only get a table of differing coefficients; for H3/H4 you get the TP equality asserted with no polynomial shown, deferred to a GitHub repo with no commit hash. That is a real reproducibility gap for a headline result. The paper also leans on [26]—an arXiv preprint—for the recursive characterization of TP, and Remark 6.1 simply asserts agreement with BKP. If that recursion or the hidden computations are wrong, Corollary 6.2 falls. It's a contained weakness: the rest of the paper doesn't depend on it, and the characteristic polynomial computations displayed for H3/H4 are checkable. But for the main advertised consequence, the evidence should be in the paper or in a versioned artifact with a commit hash and preferably independent verification.\n\nAlso minor: Prop 2.2 (planar duality) is delegated to a co-authored book chapter, and Prop 3.1 to a published paper, which is acceptable but worth noting.\n\nWho's this for? Anyone working on graph/Tutte polynomial extensions, hypergraph invariants, or hypergraph Potts models. It advances the program of a unifying hypergraph Tutte theory without claiming to finish it.\n\nRecommendation: send it to a serious referee; it deserves a proper review. But as a condition of acceptance, Section 6 needs to be made reproducible—full computational data or a versioned repository, ideally a short verification script in the paper's appendix.","headline":"Honest, solid core theory with a genuinely new deletion–contraction Tutte polynomial for hypergraphs; the headline comparison with Bernardi–Kálmán–Postnikov is real but rests on external computational and preprint dependencies that need pinning down.","tokens_in":31790,"tokens_out":2541,"would_cite":true,"duration_ms":24269,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05C31","05B35","05C65","82B20"],"pacs":[],"model":"deepseek-v4-flash","headline":"A new hypergraph Tutte polynomial supports deletion–contraction, specializes to uniform polymatroids, and proves the characteristic polynomial is not a specialization of the older polymatroid Tutte polynomial.","keywords":["hypergraph Tutte polynomial","deletion–contraction","k-polymatroids","universality","Potts model","random cluster model","distinguishing power","characteristic polynomial"],"falsifier":"Recompute $T_{HG}$ and the earlier polymatroid Tutte polynomial on the two pairs of 4-uniform hypergraphs described in Section 6 (seven- and eight-vertex examples with four and five edges) using an independent implementation of the subset-sum definitions; finding $T_P(H_1)=T_P(H_2)$ or $T_{HG}(H_3)=T_{HG}(H_4)$ would overturn the incomparability conclusion and the negative answer about the characteristic polynomial.","tokens_in":30895,"feed_emoji":"🕸️","tokens_out":8075,"duration_ms":79302,"temperature":0.7,"texified_at":"2026-08-05T21:27:40.230931+00:00","pith_summary":"The paper introduces a Tutte polynomial for hypergraphs, $T_{HG}$, whose subset-sum definition records the degree of every hyperedge. Its central achievement is a deletion–contraction recurrence that stays inside the class of hypergraphs, achieved through a degree-preserving contraction operation; earlier hypergraph Tutte extensions did not have this. The paper pairs $T_{HG}$ with a Tutte polynomial $T_k$ for $k$-polymatroids and proves that, on $(k+1)$-uniform hypergraphs, $T_{HG}$ is exactly $T_k$ on the associated polymatroid. From that bridge it derives multiplicativity, duality, a universality/recipe theorem, and a convolution formula, and shows $T_{HG}$ is equivalent to a degree-dependent hypergraph random-cluster partition function. The closing comparison shows $T_{HG}$ and a previously defined polymatroid Tutte polynomial are incomparable in distinguishing power, which answers negatively a question about whether the characteristic polynomial is a specialization of that earlier polynomial. A careful reader would care because this supplies the missing deletion–contraction backbone for hypergraph analogues of graph and matroid Tutte theory, with statistical-mechanics consequences.","texify_model":"deepseek-v4-flash","texify_usage":{"total_tokens":6588,"prompt_tokens":908,"completion_tokens":5680,"prompt_tokens_details":{"cached_tokens":0},"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":908,"completion_tokens_details":{"reasoning_tokens":4782}},"feed_headline":"A hypergraph Tutte polynomial finally has deletion–contraction","feed_subtitle":"Settles an open question: the characteristic polynomial is not a specialization of the older polymatroid Tutte polynomial.","key_machinery":"The engine is the degree data built into hyperedges. Deletion removes a hyperedge; contraction merges the vertices of a hyperedge into one vertex while keeping all other edge degrees unchanged, so the quantities $m(e)=d(e)-v(e)$ (loose multiplicity) and $\\Delta(e)=\\kappa(H\\setminus e)-\\kappa(H)$ (component drop) enter as the recursion weights $(x-1)^\\Delta$ and $(y-1)^m$. On the polymatroid side the same structure is carried by the rank function $r(A)=v(H)-\\kappa_H(A)$ and the $k$-dual $r^*(A)=k|A|+r(E\\setminus A)-r(E)$; $T_k(P)=\\sum_A (x-1)^{r(E)-r(A)}(y-1)^{k|A|-r(A)}$ is the object that supports universality and the convolution product.","core_discovery":"The central claim is that $T_{HG}(H;x,y)=\\sum_{A \\subseteq E} (x-1)^{\\kappa(A)-\\kappa(H)} (y-1)^{d(A)-|A|-v(H)+\\kappa(A)}$ is a genuine hypergraph analogue of the Tutte polynomial: for every hypergraph and every hyperedge $e$ it satisfies $T_{HG}(H) = (x-1)^{\\Delta(e)} T_{HG}(H\\setminus e) + (y-1)^{m(e)} T_{HG}(H/e)$, where $\\Delta(e) = \\kappa(H\\setminus e) - \\kappa(H)$ and $m(e) = d(e) - v(e)$. The key choice is contracting a hyperedge by identifying all of its vertices while preserving the degrees of the remaining edges; this makes the recursion close within hypergraphs. On $(k+1)$-uniform hypergraphs the degree term becomes $k|A|$ and $T_{HG}$ agrees with $T_k$ on the associated polymatroid, transferring the polymatroid universality theorem and convolution formula back to hypergraphs. The paper furthe","pith_inferences":["The paper leaves implicit that the same degree-recording device could define Tutte invariants for other multiset-based structures, such as simplicial complexes with repeated faces, where rank functions forget multiplicity.","Because THG distinguishes hypergraphs that the rank-based polymatroid polynomial does not, a natural next target is a 'degree-enriched' polymatroid Tutte polynomial that extends the earlier one without losing hyperedge degree data.","The equivalence between THG and Z_RC, together with the multivariate rank-generating-function viewpoint mentioned in the paper, suggests a testable interpolation: a multivariate THG with one edge variable per hyperedge should reduce to THG, Z_P, and the many-body Potts partition function by different substitutions, unifying the three models outside the uniform case."],"forward_implications":["Hypergraph Tutte polynomials can now be computed by deletion–contraction, and any multiplicative deletion–contraction invariant on (k+1)-uniform hypergraphs is a specialization of THG via the universality/recipe theorem.","THG supplies an organizing polynomial for the degree-dependent hypergraph random-cluster model, so evaluations and specializations of the Tutte polynomial—including its duality for planar hypergraphs—transfer to this partition function.","Because THG agrees with T_k on uniform hypergraphs, the k-polymatroid convolution formula gives hypergraph convolution identities that reduce to the classical graph Tutte convolution in the graph case.","The characteristic polynomial (and, up to a known factor, the chromatic polynomial) of a hypergraph is not in general a specialization of the earlier polymatroid Tutte polynomial; the open question is closed in the negative.","In the uniform setting, the three hypergraph partition functions (degree-dependent random cluster, degree-dependent Potts, and many-body Potts) become equivalent, so the non-uniform behavior is where the models genuinely differ."],"fun_headline_variants":["Hypergraph Tutte polynomial: deletion–contraction recursion","Tutte polynomial for hypergraphs: deletion–contraction proven","Hypergraph analogue of Tutte polynomial: deletion–contraction","New Tutte polynomial for hypergraphs with deletion–contraction","Hypergraph Tutte polynomial: deletion–contraction and consequences"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the recursion used to compute the earlier polymatroid Tutte polynomial agrees with its original defining activities, and that the two pairs of 4-uniform hypergraphs in Section 6 have exactly the polynomial equalities and inequalities asserted; these identities are checked computationally and not fully derived in the text.","fun_headline_variants_meta":{"raw":{"variants":["Hypergraph Tutte polynomial: deletion–contraction recursion","Tutte polynomial for hypergraphs: deletion–contraction proven","Hypergraph analogue of Tutte polynomial: deletion–contraction","New Tutte polynomial for hypergraphs with deletion–contraction","Hypergraph Tutte polynomial: deletion–contraction and consequences"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000767,"raw_usage":{"total_tokens":3298,"prompt_tokens":868,"completion_tokens":2430,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":2344}},"tokens_in":612,"tokens_out":2430,"duration_ms":17552,"temperature":1.0,"reasoning_tokens":2344,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T01:25:03.333260+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $T_{HG}$ and the earlier polymatroid Tutte polynomial on the two pairs of 4-uniform hypergraphs described in Section 6 (seven- and eight-vertex examples with four and five edges) using an independent implementation of the subset-sum definitions; finding $T_P(H_1)=T_P(H_2)$ or $T_{HG}(H_3)=T_{HG}(H_4)$ would overturn the incomparability conclusion and the negative answer about the characteristic polynomial.","supporting_citations":[],"review_version":1}