{"id":"840ae394-5d8b-48b7-b81b-9b256f3fbca0","arxiv_id":"2607.23890","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.5,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Stratified motivic invariants of M̄0,n and Y[n] admit explicit inductive formulas via echelon towers, and the bivariate deformations of even Poincaré polynomials in prior work are precisely those stratified virtual Poincaré polynomials.","lead":"The paper defines stratified motivic invariants for varieties with natural filtrations and proves inductive formulas for them on moduli of genus-0 stable curves and Fulton–MacPherson spaces. It identifies the mysterious bivariate polynomials from an earlier real-rootedness proof as exactly these stratified virtual Poincaré polynomials.","discovery_kind":"unification","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"No significant objection identified. The load-bearing fiber computations (Prop. 3.1, eqs. (18)–(19)) and the algebra of (13)/(21) independently verify; the only residual dependence is the precisely stated recurrence match with the companion paper [BK26].","rationale":"The reader identified solidity of the echelon towers — the motivic local triviality and constant-fiber-invariant claims in Prop. 3.1 and eqs. (18)–(19) — as the weakest assumption, and that is indeed the right place to look: every downstream result (the recurrences (13)/(21), the PDE (24), the recovery of Manin's formulas, and the identification with the [BK26] deformations) is formal algebra once the fiber invariants are known. However, on inspection the fiber computations are not merely asserted; they are derived from explicit geometric descriptions, and I verified them arithmetically (node/mark counts on genus-0 trees; blowup-and-glue motivic bookkeeping for Y′) and against independent data (known Poincaré polynomials of M̄0,4 and M̄0,5, the blowup construction of Y[2] and Y[3], and the x²-coefficient of Manin's generating function). All agree. The remaining dependence on the companion paper is confined to the exact statement of [BK26, Def. 2.2], which is a verbatim-checkable external fact rather than an internal gap, and the reader's circularity assessment (limited to the application, not the identification) is accurate. Given fully written elementary arguments, multiple independent consistency checks that pass, and a falsifiable coefficient-level prediction (15), the ACCEPT with high confidence is warranted and no adjustment is needed.","tokens_in":10688,"tokens_out":10897,"duration_ms":393972,"concrete_test":"Two quick independent checks would settle the residual risk completely: (a) open [BK26, Definition 2.2] and confirm its recurrence and initial condition coincide with (14) under L ↦ t² and F_2(y) = y — if either differs, Corollary 3.3 fails outright; (b) compute F_4(y) from (14) and verify [y^{j+1}]F_4(y, t²) for j = 0,1,2 against the virtual Poincaré polynomials of the three strata of M̄0,5 computed directly from the dual-graph stratification (expected: (L−2)(L−3), 10(L−2), 15 with L = t²); repeat for P¹[2] (expected F = (L+y)(L+1)). Any mismatch localizes an error in the fiber counts of Prop. 3.1 or eqs. (18)–(19).","verdict_should_be":"UNCHANGED","load_bearing_attack":"I re-derived the steps on which everything rests and could not find a soft spot that survives checking. (1) In Prop. 3.1, the fiber of πn,j,0 over C ∈ M0,n,j is C minus its n+j special points (n marks + j nodes, nodes counted once as points of C): H = (j+1)L + 1 − (n+j) = (j+1)L − j − n + 1, matching the paper. The fiber of πn,j,1 is the finite set of n+j−1 special points of the (j−1)-stratum curve (bubbling at a mark or node is rigid since M0,3 = pt), matching B_{n,j} = n+j−1. (2) Equation (12) → (13) is correct term-by-term, and the stated examples verify: F_{M0,4} = (L−2)+3y and F_{M0,5} = (L−2)(L−3)+(10L−20)y+15y² reproduce, at y=1, the known Poincaré polynomials of M̄0,4 ≅ P¹ and of the degree-5 del Pezzo M̄0,5 (L²+5L+1). Moreover the coefficients match direct stratum counts: open stratum (L−2)(L−3), ten boundary divisors each with interior M0,4 giving 10(L−2), and 15 codimension-2 points — exactly as (15) predicts. (3) In the FM case, H_{Y′} = H_Y + j(H_{P^m}−1) by the blowup-and-glue count, and removing the n marks and j disjoint intersection P^{m−1}'s gives H_Y − n + j(L^m − 1), matching (18); the (19) count of n+j−1 copies of P^{m−1} (one per leg or edge) is consistent with the standard C*-quotient description of the bubbling direction. (4) The low-n checks for (21) against blowup constructions (Y[2] = bl_∆(Y²); the n=2 example at y=1) all confirm, and the final identity (31) expanded to order x² yields H_Y(H_Y − 1 + H_{P^{m−1}}) = H_{Y[2]}, agreeing with the blowup formula. The one thing not internally verifiable from this manuscript is that [BK26, Def. 2.2] is literally (14) with L = t² and initial condition F_2(y) = y; but that is a precise, checkable external statement, not a hidden assumption, and the reader already routed the companion-paper dependence correctly.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The paper introduces, for a stratified quasi-projective variety X and a motivic invariant H, the stratified invariant F_X(y) = Σ y^j H_{X°_j}, which interpolates between H_{X°} (y=0) and H_X (y=1). It axiomatizes families of stratified moduli spaces connected by forgetful morphisms as \"echelon towers\" (Definition 2.3), with a \"solidity\" condition (motivic local triviality of the two restricted maps π_{n,j,0}, π_{n,j,1}) that yields the inductive identity (7). The two main theorems compute the fiber invariants explicitly for the moduli spaces M̄_{0,n} (Proposition 3.1, Eq. (12)) and the Fulton–MacPherson compactifications Y[n] (Eqs. (18)–(20)), producing differential recursions (13) and (21). As an application, the bivariate polynomials F_n(y,t) and F̃_n(y,t) introduced ad hoc in the companion paper [BK26] to prove real-rootedness of the even-degree Poincaré polynomials of M̄_{0,n+1} and P^1[n] are identified with stratified virtual Poincaré polynomials (Corollaries 3.3 and 4.4). A characteristic-method solution of the generating-function PDE (24) yields a new elementary proof of Manin's formulas for H_{Y[n]} (Theorem 4.5).","tokens_in":11216,"tokens_out":11461,"duration_ms":87701,"significance":"The paper gives a genuine geometric interpretation of the bivariate deformation that appeared \"out of nowhere\" in the AI-assisted real-rootedness proof of [BK26]: the coefficients of F_n(y,t²) are exactly the virtual Poincaré polynomials of the dual-graph strata (Eq. (15)). This is a satisfying and non-obvious demystification. Independently of the companion paper, the recursions (13) and (21) are effective, graph-sum-free computational tools; the paper earns credibility by reproducing known results in checkable cases — F_{M̄0,5} at y=1 gives the degree-5 del Pezzo polynomial L²+5L+1 with the correct stratum-by-stratum count (10(L−2) from ten boundary divisors, 15 from codimension-2 strata), the n=2 Fulton–MacPherson example matches the blowup Y[2] = bl_Δ(Y²), and specialization to y=1 recovers Manin's formulas with a fresh characteristic-curve proof. I independently verified the load-bearing computations: the fiber class (j+1)L−j−n+1 in Proposition 3.1 equals H_C − (n+j) with H_C = (j+1)(L+1)−j; Eq. (18) follows from H_{Y′} = H_Y + j(H_{P^m}−1); the characteristic computation (26)–(31) is correct line by line. The results are correct as far as I can check and should be of interest","major_comments":[{"comment":"Solidity — the motivic local triviality of π_{n,j,0} and π_{n,j,1} — is the hypothesis that makes (7), and hence the recursions (13) and (21), valid; it is therefore load-bearing for both main theorems. However, its verification is compressed to one sentence in each case ('by (8), we can decompose (11) ... and find that the morphisms in (11) are motivically locally trivial'; similarly after (17)). The fiber motivic invariants are computed explicitly and check out (I verified all of (12), (18), (19) independently), but local triviality itself is not argued. For M̄_{0,n} one expects: over each stratum M_{0,n,γ} the restriction of the universal curve with its special sections removed is Zariski locally trivial because, after normalization, the family is obtained by gluing trivial families of pointed P¹'s, and the attaching sections remain disjoint within the stratum. For (19), one needs tha","section":"§2.2, Definition 2.3 and its use in Proposition 3.1 and §4 (Eqs. (17)-(20))"}],"minor_comments":[{"comment":"The formula for F_{Y[3]} contains H_{P^{d−1}} (twice), with d undefined; this should be H_{P^{m−1}}.","section":"§4, Example after (22)"},{"comment":"The exponent 'Lm−1' is typographically ambiguous between L^{m}−1 and L^{m−1}. Solving the characteristic equation myself confirms the intended meaning is L^{m}−1 (e.g., (28) reads y/(y+L−1) = (1/L)(z/u)^{L^m−1}); please add parentheses to remove the ambiguity.","section":"§4, Eqs. (28), (30), (31)"},{"comment":"Notation oscillates between M_{0,n,j} and M̄_{0,n,j} for the same locally closed strata; since M_{0,n,γ} denotes the open locus inside M̄_{0,n}, a uniform convention would help.","section":"§3, Eq. (10) and Proposition 3.1"},{"comment":"The assertion that M̄_{g,n} for g≥1 forms a non-solid echelon tower would benefit from one line of explanation: the fiber of π_{n,j,0} over C is still C minus special points with constant motivic class, so the failure is genuinely in Zariski local triviality of the universal family over strata (smooth higher-genus curves have moduli). As stated, a reader might mistakenly think the fiber invariant itself fails to be constant.","section":"Remark 3.4"},{"comment":"In Corollary 3.3 the identification F_n(y,t²) = yF_{M̄0,n+1}(y) implicitly uses R = Z[t] with L = t²; stating this substitution explicitly (it is only explained via P^even_n(t²) = P_{M̄0,n}(t)) would make the corollary self-contained.","section":"§1 and Corollary 3.3"},{"comment":"§4: 'autormorphism group' (twice); §4 after (19): 'the union of of n+j−1 copies'; §2.2: 'motivially locally trivialif it factors'. Also [BK26] should include the arXiv identifier in the bibliography as it currently does; fine, but 'poincaré' is mis-capitalized in the [BK26] title.","section":"Typos"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is a companion to [BK26] by the same authors, and its headline application (Corollaries 3.3 and 4.4) is the reinterpretation of polynomials introduced there. The recursion results (Theorems 3.2, 4.1) and the new proof of Manin's formulas stand independently, so the paper does not feel like salami slicing, but the editor should be aware of the tight coupling: the significance of the \"mystery\" being resolved is inherited from the companion paper, which is itself not yet refereed. The mathematics I could check is correct, and my one substantive request (a proof of solidity) is presentational rather than a gap I expect to fail."},"author_rebuttal":{"model":"moonshotai/kimi-k3","summary":"We thank the referee for the careful and engaged report, for the positive assessment of the results, and especially for independently verifying the load-bearing computations (Proposition 3.1, Eq. (18), and the characteristic computation (26)–(31)). We are glad the referee finds the geometric interpretation of the bivariate deformations of [BK26] satisfying. The referee's sole major comment concerns the verification of solidity, which is indeed load-bearing for both main theorems and is currently compressed to one sentence in each case. We agree this deserves a fuller treatment and will expand the arguments in the revision, as detailed below.","responses":[{"response":"We agree entirely. Solidity is the hypothesis that makes identity (7), and therefore both recursions, valid, so it should be argued rather than asserted. In the revision we will add a dedicated lemma in each section giving the full local triviality argument along exactly the lines the referee sketches. For M̄_{0,n}: over the stratum M_{0,n,γ} the restriction of the universal stable curve, after normalization, decomposes as a disjoint union of constant families of pointed P^1's (one per vertex of γ), since the moduli of a k(v)-pointed P^1 are rigid once the tree structure is fixed; the attaching sections stay disjoint within the stratum because the dual graph is constant, so π_{n,j,0} is Zariski locally trivial with fiber a pointed P^1-tree minus its n+j special sections, and π_{n,j,1} is locally trivial with fiber the finite set of special points (n+j−1 of them). For Y[n]: over each stratum Y[n]°_γ the family of degenerate varieties Y′ is locally trivial because γ fixes the blowup tree and the blowup centers and gluing loci vary in disjoint families; the fiber of π_{n,j,0} is the complement of the marked points and the P^{m−1}-intersections, and the fiber of π_{n,j,1} is the union of n+j−1 copies of P^{m−1} arising as C*-quotients as described before (19). We will also spell out explicitly why the complement description gives a motivic (not merely Zariski) local trivialization in the sense of Definition 2.3. The fiber invariant computations (12), (18), (19), which the referee verified, remain unchanged; the revision adds the missing triviality argument.","revision_made":"yes","referee_comment":"Solidity (motivic local triviality of pi_{n,j,0} and pi_{n,j,1}) is load-bearing for (7) and hence for (13) and (21), but its verification is compressed to one sentence in each case. The fiber invariants are computed explicitly and check out, but local triviality itself is not argued. For M̄_{0,n} one expects an argument via normalization: over each stratum the universal curve is glued from trivial families of pointed P^1's, and the attaching sections remain disjoint within the stratum. For (19) one needs a similar argument."}],"tokens_in":10675,"tokens_out":761,"duration_ms":11366,"standing_objections":[]},"desk_editor":{"model":"grok-4.5","letter":"The punchline is simple: the mysterious Fn(y,t) from the companion real-rootedness paper is just y times the stratified virtual Poincaré polynomial of M̄0,n+1, and the same for P1[n]. They get there by introducing stratified motivic invariants FX(y) and solid echelon towers, then deriving clean inductive formulas (13) and (21).\n\nWhat is new is the organizational language and the recursions themselves. Once you accept that the forgetful maps restrict to motivically locally trivial fibrations on the open strata with constant fiber invariants, the passage from the fiber formulas (12) and (20) to the differential recursions is elementary calculus and is written out in full. The fiber counts check: for M̄0,n the πn,j,0 fiber is the curve minus its n+j special points, giving (j+1)L-j-n+1; the πn,j,1 fiber is a finite set of size n+j-1. The FM case is the same style of blow-up-and-glue arithmetic. Low-n specializations recover the known Poincaré polynomials of M̄0,4 ≅ P1 and M̄0,5, and the y=1 generating function recovers Manin’s formulas by the method of characteristics. That is real, checkable work.\n\nThe soft spots are minor and local. The “solid” condition is verified case-by-case rather than by a general theorem; if someone wants to push the same formalism to Mg,n or stable maps they will have to do the fiber analysis again (the paper already notes those towers are not solid). Dependence on the companion is only for the name of Fn and the real-rootedness application; the identification itself is independent. No free parameters, no circularity in the main claims.\n\nThis is for people who already care about Poincaré polynomials or motivic invariants of classical moduli spaces. It is short, algebraic, and fully explicit. I would send it to a serious referee without hesitation; the math is solid enough that the only useful referee comments will be about exposition and possible extensions. Engage with it if you work in this area.","headline":"Clean geometric meaning for the BK26 bivariate polynomials via new inductive formulas for stratified motivic invariants of M̄0,n and Y[n].","tokens_in":12379,"tokens_out":524,"would_cite":true,"duration_ms":9429,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H10","14N20","14C17","05A15"],"pacs":[],"model":"grok-4.5","headline":"The mysterious bivariate polynomials that proved real-rootedness for M̄0,n are stratified virtual Poincaré polynomials.","keywords":["stratified motivic invariant","echelon tower","moduli of stable curves","Fulton-MacPherson compactification","virtual Poincaré polynomial","bivariate deformation","real-rootedness"],"falsifier":"Compute the stratified virtual Poincaré polynomial of M̄0,6 or P1[4] both from the inductive formula and by direct enumeration of dual-graph strata; any mismatch of coefficients falsifies the claim that the earlier bivariate polynomials coincide with the stratified invariants.","tokens_in":11967,"feed_emoji":"📐","tokens_out":981,"duration_ms":18977,"temperature":0.7,"pith_summary":"This paper gives a geometric meaning to a pair of bivariate polynomials that earlier work used to prove that the even-degree Betti numbers of the moduli spaces of stable genus-zero curves and of the Fulton–MacPherson compactifications of configurations on the line form ultra-log-concave sequences. The authors introduce stratified motivic invariants that interpolate between the ordinary invariant of a variety and the invariant of its open interior, then codify a recursive structure (an “echelon tower”) satisfied by many moduli spaces under forgetful maps. For the spaces M̄0,n and Y[n] the towers are solid, yielding clean inductive formulae for the stratified invariants. Specialising to the virtual Poincaré polynomial recovers exactly the earlier bivariate deformations, so those polynomials are no longer mysterious: their coefficients are the virtual Poincaré polynomials of the dual-graph strata. The same formulae also recover classical generating-function expressions for the ordinary Poincaré polynomials by elementary calculus.","feed_headline":"Mysterious bivariate polynomials are stratified Poincaré invariants","feed_subtitle":"Inductive formulae from echelon towers identify the deformations that proved real-rootedness of M̄0,n","key_machinery":"A solid echelon tower: a sequence of stratified varieties X[n] linked by morphisms πn that send the open codimension-j stratum into the union of the codimension-j and codimension-(j-1) strata of X[n], with the two restricted maps motivically locally trivial and of constant fibre invariant. The constant-fibre condition converts the stratification into a linear recurrence for the generating function FX[n](y).","core_discovery":"The bivariate deformations Fn(y,t) and eFn(y,t) of the even-degree Poincaré polynomials of M̄0,n+1 and P1[n] that appeared in earlier work are precisely the stratified virtual Poincaré polynomials of those spaces (up to the elementary factor y in the first case). Both families of stratified invariants are uniquely determined by explicit first-order inductive formulae coming from solid echelon towers.","pith_inferences":["If the solid-echelon condition can be verified for other compactified configuration spaces or for Hassett spaces, the same inductive machine would immediately produce stratified invariants and candidate real-rooted polynomials.","The failure of solidity for Mg,n (g>0) and for stable-map spaces suggests that a weaker “virtual-fibre” version of the tower might still control generating functions in those settings.","Once the coefficients are known to be stratum-wise virtual Poincaré polynomials, positivity or unimodality questions about those strata become equivalent to coefficient-wise questions about Fn."],"forward_implications":["The coefficients of the earlier bivariate polynomials are now identified as virtual Poincaré polynomials of explicit strata, giving them independent geometric content.","Specialising the inductive formulae at y=1 recovers all previously known generating functions for the ordinary Poincaré polynomials of M̄0,n and Y[n] by elementary calculus.","The same echelon-tower formalism applies verbatim to any other sequence of moduli spaces whose forgetful maps satisfy the solid-fibre condition.","The paper notes that the ideas may extend to Chow rings of loopless matroids and their Poincaré polynomials."],"fun_headline_variants":["Bivariate Poincaré deformations are stratified virtual invariants","Echelon towers give inductive formulae for stratified Poincaré polynomials","Stratified virtual Poincaré polynomials match the bivariate deformations","Fn(y,t) and eFn(y,t) equal stratified virtual Poincaré polynomials","Inductive echelon formulae identify the bivariate Poincaré deformations"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The restricted forgetful maps on open dual-graph strata must be motivically locally trivial with fibres of constant motivic invariant; without that constancy the inductive formulae fail.","fun_headline_variants_meta":{"raw":{"variants":["Bivariate Poincaré deformations are stratified virtual invariants","Echelon towers give inductive formulae for stratified Poincaré polynomials","Stratified virtual Poincaré polynomials match the bivariate deformations","Fn(y,t) and eFn(y,t) equal stratified virtual Poincaré polynomials","Inductive echelon formulae identify the bivariate Poincaré deformations"]},"model":"grok-4.5","effort":"low","cost_usd":0.004488,"raw_usage":{"total_tokens":1255,"prompt_tokens":706,"num_sources_used":0,"completion_tokens":65,"cost_in_usd_ticks":44884000,"prompt_tokens_details":{"text_tokens":706,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":484,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":706,"tokens_out":65,"duration_ms":9187,"temperature":1.0,"reasoning_tokens":484,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T14:40:13.810592+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Compute the stratified virtual Poincaré polynomial of M̄0,6 or P1[4] both from the inductive formula and by direct enumeration of dual-graph strata; any mismatch of coefficients falsifies the claim that the earlier bivariate polynomials coincide with the stratified invariants.","supporting_citations":[],"review_version":2}