{"id":"b85858e2-ff00-4e07-bddf-30d5fb4c9bf6","arxiv_id":"2508.15538","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Chow and augmented Chow polynomials of finite graded simplicial posets with nonnegative h-vectors, including Cohen-Macaulay simplicial posets, are real-rooted, with an interlacing result for duals.","lead":"This paper proves a structural result in combinatorics: certain polynomials associated to a broad class of posets always have only real roots. It confirms and extends a known conjecture to a substantial family of examples, with a proof based on interlacing roots.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the interlacing induction is sound once obvious index typos in Lemmas 3.1/3.8 are corrected.","rationale":"The reader's weakest-assumption identification pointed at the induction verification around Lemma 3.8 and condition (3.7). I agree that this is the most delicate part of the paper, but after re-deriving the recursion and checking the claimed interlacing steps, the intended induction goes through. The apparent issues are typographical: Lemma 3.1 conflates the Chow polynomial of the dual with the augmented Chow polynomial of the dual, and Lemma 3.8 states condition (3.7) with an index that makes it vacuous; the actual nontrivial diagonal interlacing is correctly proved in Lemma 3.13. These typos are concerning for exposition but not for the central mathematical claim. The key ingredients—Equation (1.3), the recursive description of the p-polynomials, Lemma 2.3, and the five-part induction in Lemma 3.8—are standard and consistent after the evident corrections. I therefore find no load-bearing objection and would leave the reader's ACCEPT verdict unchanged.","tokens_in":12791,"tokens_out":33218,"duration_ms":306264,"concrete_test":"Use exact arithmetic to verify Theorem 3.3 up to n=8: generate p^{S⊆T}_{n,k} from the recursion (3.1), compute real roots (or Sturm sequences), and test interlacing along every row, column, and the diagonal edge p^{T\\{1}}_{n,n-1} ⪯ p^{{1}⊆T}_{n,1} for T=[1,n] and [1,n-1]. A failure would pinpoint the induction step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I could not find a load-bearing flaw in the central claim. The proof hinges on Theorem 3.3, and the induction through Lemmas 3.9–3.13 is valid in substance. The printed text has local errors: Lemma 3.1's second display should refer to γ(H_{P^*}), not γ(H^aug_{P^*}) (the augmented equality is the third display); Lemma 3.8's (3.7) is written as p^{T\\{1}}_{n,n-1} ⪯ p^{{1}⊆T}_{n,0}, which is a zero polynomial, whereas the edge used in the zero-omitted diagram and proved in Lemma 3.13 is p^{T\\{1}}_{n,n-1} ⪯ p^{{1}⊆T}_{n,1}; Lemmas 3.9/3.10 have off-by-one T-ranges. These are evident from context and do not affect the argument after correction. No unsupported external assumption beyond the standard flag-h identity (1.3) is required.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Theorem 1.1: for every finite graded simplicial poset P with positive h-vector, the Chow and augmented Chow polynomials of \\hat P are real-rooted; Theorem 1.2 adds that for the dual poset \\hat P^*, the Chow polynomial is real-rooted and its roots interlace those of the augmented Chow polynomial. The proof rewrites the flag h-vector via identity (1.3), expresses the γ-polynomials of Chow polynomials as nonnegative combinations of a family of polynomials p^{S⊆T}_{n,k}, and establishes an interlacing diagram (Theorem 3.3) by induction with explicit base cases. The main theorem implies the FMV24 conjecture for Cohen–Macaulay simplicial posets and recovers previously known cases for uniform matroids.","tokens_in":13077,"tokens_out":16411,"duration_ms":162697,"significance":"The result is significant: it proves a conjecture for a natural class of posets and introduces an interlacing machinery that also gives a new proof of real-rootedness of Eulerian and binomial Eulerian polynomials. The argument is largely self-contained, does not depend on the conjectures under discussion, and the finite base cases are explicit. The positive-h-vector assumption is clearly stated and used exactly where needed, so the scope and limitations of the method are transparent. The decomposition of Chow polynomials into interlacing palindromic components is elegant and likely to be useful beyond this paper.","major_comments":[{"comment":"The second display of Lemma 3.1 reads γ(H^{aug}_{\\hat P^*}; x) = ∑_{k=0}^n h_k p^{[2,n]}_{n,k}(x), but the third display asserts γ(H^{aug}_{\\hat P}; x) = γ(H^{aug}_{\\hat P^*}; x) = ∑_{k=0}^n h_k p^{[1,n]}_{n,k}(x). These cannot both be correct. The intended statement, used in the proof of Theorems 1.1/1.2, is γ(H_{\\hat P^*}; x) = ∑_{k=0}^n h_k p^{[2,n]}_{n,k}(x). As printed, the proof of Theorem 1.2 does not follow from the stated lemma.","section":"Lemma 3.1"},{"comment":"The displayed sums in this proof start at k=1, reading ∑_{k=1}^n h_k p^T_{n,k}, and are then equated with γ-polynomials that Lemma 3.1 expresses as sums from k=0. The omitted h_0 p^T_{n,0} term is generally nonzero; Example 3.2 gives γ(H_{\\hat P}; x)=1+5x while ∑_{k=1}^2 h_k p^{{1}}_{2,k}=5x. The index k=1 should be k=0 throughout this passage. As printed, the main theorem is not established by the displayed equalities.","section":"Proof of Theorems 1.1/1.2"},{"comment":"Condition (3.7) is stated as p^{T\\{1}}_{n,n-1} ⪯ p^{{1}⊆T}_{n,0}, but the same lemma says p^{{1}⊆T}_{n,0} ≡ 0, so the condition is vacuous. The zero-omitted interlacing diagram requires the edge p^{T\\{1}}_{n,n-1} ⪯ p^{{1}⊆T}_{n,1}, which is exactly what Lemma 3.13 proves. The subscript n,0 should be n,1 in (3.7), and the proof of Lemma 3.8 should refer to the nonzero bottom-left entry of the zero-omitted diagram.","section":"Lemma 3.8, Eq. (3.7)"},{"comment":"The index ranges in these induction lemmas do not match the cases needed for Theorem 3.3. For D_n([1,n]) the top, middle, and bottom rows involve T = [2,n], [1,n], and {1}⊆[1,n], respectively; the lemmas state ranges such as T = [2,n−1] or [2,n−2] in Lemma 3.9 and T = [1,n−2] or [1,n−1] in Lemma 3.10, and Lemma 3.11 states p^T_{n,k} instead of p^{{1}⊆T}_{n,k}. These are evidently index slips, but as printed the induction step for D_n([1,n]) is not literally covered. The statements and proofs should be adjusted to the correct ranges and symbols.","section":"Lemmas 3.9–3.11"}],"minor_comments":[{"comment":"The definition of S−1 and T−1 is easy to misread: writing S−1 = {i−1 | i∈S}\\{1} seems to remove the shifted element 1, whereas the intended meaning, used in the examples, is to remove 1 from S before shifting. Please clarify, e.g. define S−1 = {i−1 | i∈S, i>1}.","section":"Section 3, recursion after (3.1)"},{"comment":"The proof says the two base cases for n=1 are given in Lemmas 3.6 and 3.7, but those lemmas treat D_2({1}) and D_2({1,2}), i.e. n=2. The wording should be adjusted.","section":"Proof of Theorem 3.3"},{"comment":"The statement of this lemma repeats the notation of Lemma 3.10. Given its title and proof, it should assert interlacing for the bottom row p^{{1}⊆T}_{n,k}, not for p^T_{n,k}.","section":"Lemma 3.11"},{"comment":"The remark asserts a nontrivial generalization of Theorem 3.3 with the proof omitted because it is not needed. Since the remark is not used, either label it as a conjecture/planned result or include the proof, otherwise the statement is unverifiable.","section":"Remark 3.4"},{"comment":"The proof refers to a 'pink diagonal arrow' in Theorem 3.3. Color-dependent references are not robust in print or grayscale; describe the arrow by its endpoints.","section":"Proof of Theorems 1.1/1.2"}],"recommendation":"major_revision","confidential_remarks":"I believe the main theorem is correct and the argument is essentially sound, but the number of indexing errors at load-bearing points is high: the second display of Lemma 3.1 names the wrong polynomial, the proof of the main theorems sums from k=1 instead of k=0, condition (3.7) in Lemma 3.8 has the wrong endpoint, and the ranges in Lemmas 3.9–3.11 do not cover the required cases. These are all local and seem easily correctable, so I recommend major revision rather than rejection. After the authors fix these statements and re-verify the induction, I would expect the paper to be acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth knowing: this paper does what it says. For finite graded simplicial posets with positive h-vector, the Chow and augmented Chow polynomials are real-rooted, and for the dual poset the Chow interlaces the augmented. That settles the Ferroni--Matherne--Vecchi conjecture for the simplicial Cohen-Macaulay case, and it recovers the uniform matroid results of FMSV24 and BV25.\n\nWhat is actually new is the proof mechanism. The h-vector expansion (1.3) rewrites the flag h-vector in terms of descent sets of permutations, and the interlacing diagram in Theorem 3.3 is the engine. That theorem is a careful induction with five conditions, and it gives a new decomposition of Eulerian and binomial Eulerian polynomials into interlacing palindromic pieces. The paper is self-contained: the interlacing facts in Section 2 are standard, and the main theorems follow cleanly once Theorem 3.3 is in place.\n\nThe soft spots are minor. The stress-test found a few local index typos: in Lemma 3.1, the second display should refer to gamma(H_{P^*}), not gamma(H^aug_{P^*}); in Lemma 3.8, condition (3.7) should involve p^{{1}⊆T}_{n,1}, not p^{{1}⊆T}_{n,0}; and Lemmas 3.9/3.10 have off-by-one T-ranges. These are clear from context and do not affect the substance. I could not find a load-bearing flaw in the main induction, and the proof of Theorem 3.3 is robust enough that I believe it after reading it.\n\nWho should read this: anyone working on Chow polynomials of matroids or posets, real-rootedness, or gamma-positivity. It is a serious paper that deserves referee time. I would send it to a good combinatorics journal.","headline":"Settles a conjecture for simplicial Cohen-Macaulay posets with a genuinely new interlacing argument; the local typos are cosmetic.","tokens_in":13520,"tokens_out":2555,"would_cite":true,"duration_ms":24968,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05A20","06A07","05A15"],"pacs":[],"model":"deepseek-v4-flash","headline":"Simplicial posets with positive h-vector have real-rooted Chow and augmented Chow polynomials.","keywords":["Chow polynomial","simplicial poset","real-rooted","h-vector","interlacing","gamma-expansion","Cohen-Macaulay","flag h-vector"],"falsifier":"Compute the Chow polynomial $H_{\\hat{P}}(x)$ of any finite graded simplicial poset with nonnegative $h$-vector—for instance the lattice of flats of the uniform matroid $U_{5,8}$ with its top element attached—and test every root for being real; a single non-real root would contradict Theorem 1.1.","tokens_in":12741,"feed_emoji":"🧮","tokens_out":7023,"duration_ms":66698,"temperature":0.7,"texified_at":"2026-08-05T19:52:26.884563+00:00","pith_summary":"The paper proves a structural theorem in enumerative combinatorics: if a finite graded simplicial poset has an $h$-vector with no negative entries, then the Chow polynomial and augmented Chow polynomial of the poset with a top element added are real-rooted, meaning all their roots lie on the real line. Real-rootedness is the strongest common form of log-concavity for such polynomials, and it is what many positivity conjectures in matroid theory reduce to. The result settles the known conjecture on Cohen–Macaulay simplicial posets and their duals, and it recovers, in one stroke, previously proved real-rootedness results for uniform matroids. A sympathetic reader should care because the proof introduces a family of permutation-descent polynomials whose interlacing structure is flexible enough to yield the result from the $h$-vector alone.","texify_model":"deepseek-v4-flash","texify_usage":{"total_tokens":8939,"prompt_tokens":765,"completion_tokens":8174,"prompt_tokens_details":{"cached_tokens":0},"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":765,"completion_tokens_details":{"reasoning_tokens":7453}},"feed_headline":"Positive h-vectors force real-rooted Chow polynomials","feed_subtitle":"A permutation-descent interlacing argument turns nonnegative h-vectors into root-location guarantees for Chow polynomials.","key_machinery":"The load-bearing object is the family $p^T_{n,k}(x)$ of polynomials that count permutations $w\\in S_{n+1}$ with $w(1)=k+1$ and descent set an isolated set inside $T$. Equation (1.3) expresses the flag $h$-vector $\\beta$ of a simplicial poset through these counts, making each Chow $\\gamma$-polynomial a nonnegative linear combination of one row of the interlacing diagram. Theorem 3.3 asserts that, for $T=[1,n]$ or $T=[1,n-1]$, the three-row diagram built from $p^{T\\setminus\\{1\\}}_{n,k}$, $p^T_{n,k}$, $p^{\\{1\\}\\subseteq T}_{n,k}$ is interlacing; this is what converts a coordinate-wise positivity assumption (the $h$-vector) into a root-location conclusion.","core_discovery":"The central claim is Theorem 1.1: for a finite graded simplicial poset $P$ with positive $h$-vector, both $H_{\\hat{P}}(x)$ and $H^{\\mathrm{aug}}_{\\hat{P}}(x)$ are real-rooted. The proof works by rewriting the flag $h$-vector of $\\hat{P}$ with the identity relating $\\beta$ to the $h$-vector and to permutation descent sets, so that each $\\gamma$-polynomial of the Chow-type polynomials becomes a nonnegative combination of polynomials $p^T_{n,k}$ counting permutations with prescribed descent sets. The technical heart is Theorem 3.3, an interlacing diagram for these $p$-polynomials: every path through the diagram is an interlacing sequence, and hence every nonnegative linear combination is real-rooted. Passing from $\\gamma$-polynomials back to the or","pith_inferences":["The paper leaves open whether real-rootedness persists without h-positivity; the conjecture in the paper suggests one should search for a simplicial poset with a negative h-vector entry whose Chow polynomial still has only real roots.","Because the proof is purely a property of permutation-descent interlacing, one can test whether analogous interlacing diagrams exist for other flag-like invariants of matroids, which would yield real-rootedness by the same mechanism.","A natural extension is to replace the nonnegative h-vector assumption with log-concavity or gamma-positivity of the h-vector; the interlacing machinery would remain applicable as long as the coefficients in the linear combination stay nonnegative."],"forward_implications":["The conjecture that Cohen–Macaulay posets have real-rooted Chow polynomials is proved in the simplicial case, since Cohen–Macaulay simplicial posets are h-positive.","The dual variant gives an interlacing statement between the Chow polynomial of a dual poset and the augmented Chow polynomial of the original, for this class.","For uniform matroids, the theorem recovers real-rootedness of both Chow and augmented Chow polynomials as special cases.","For the Boolean lattice, the decomposition of the Eulerian and binomial Eulerian polynomials into interlacing palindromic summands yields an alternative proof of their real-rootedness.","The same interlacing diagram supplies real-rootedness for any nonnegative h-vector, not just those coming from Cohen–Macaulay posets."],"supporting_citations":[{"why":"Supplies the definition of the h-vector and the theorem that Cohen–Macaulay posets are h-positive.","marker":"[Sta96]"},{"why":"Defines the Chow and augmented Chow polynomials and states the conjecture the main theorem settles.","marker":"[FMV24]"},{"why":"Provides the companion result for chain polynomials and the same initial rewriting of the flag h-vector under h-positivity.","marker":"[ADKE24]"},{"why":"Gives the characterization of interlacing sequences used to turn adjacent interlacing into full interlacing along paths.","marker":"[Wag92]"},{"why":"Supplies the closure properties of interlacing sequences that drive the induction, especially Lemma 2.3.","marker":"[Br¨ a15]"},{"why":"Establishes that a palindromic polynomial is real-rooted exactly when its γ-polynomial is, the bridge used in both directions.","marker":"[Pet15]"},{"why":"Proves real-rootedness of augmented Chow polynomials for uniform matroids, a case recovered here.","marker":"[FMSV24]"},{"why":"Proves real-rootedness of Chow polynomials for uniform matroids, another recovered case.","marker":"[BV25]"},{"why":"Identifies the Chow polynomial of a Boolean lattice with the Eulerian polynomial, giving the Eulerian corollary.","marker":"[Ham17]"},{"why":"Identifies the augmented Chow polynomial of a Boolean lattice with the binomial Eulerian polynomial, giving the binomial Eulerian corollary.","marker":"[EHL23]"}],"fun_headline_variants":["Real-rooted Chow polynomials from positive h-vectors","h-vector positivity forces real-rooted Chow polynomials","Interlacing argument for real-rooted Chow from positive h-vectors","When h-vectors are positive, Chow polynomials are real-rooted"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the $h$-vector has no negative entries; without it, the proof's nonnegative linear combinations of interlacing polynomials are not guaranteed real-rooted.","fun_headline_variants_meta":{"raw":{"variants":["Real-rooted Chow polynomials from positive h-vectors","h-vector positivity forces real-rooted Chow polynomials","Interlacing argument for real-rooted Chow from positive h-vectors","When h-vectors are positive, Chow polynomials are real-rooted"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000771,"raw_usage":{"total_tokens":3174,"prompt_tokens":589,"completion_tokens":2585,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":333,"completion_tokens_details":{"reasoning_tokens":2527}},"tokens_in":333,"tokens_out":2585,"duration_ms":21313,"temperature":1.0,"reasoning_tokens":2527,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T17:48:15.386930+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Chow polynomial $H_{\\hat{P}}(x)$ of any finite graded simplicial poset with nonnegative $h$-vector—for instance the lattice of flats of the uniform matroid $U_{5,8}$ with its top element attached—and test every root for being real; a single non-real root would contradict Theorem 1.1.","supporting_citations":[],"review_version":1}