{"id":"2e406469-570b-47ed-948c-709aac0092f1","arxiv_id":"2607.06477","paper_version":1,"verdict":"ACCEPT","confidence":"UNKNOWN","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"A generalised cross-ratio of four planes in projective n-space is shown to equal the augmented height pairing of the corresponding cycles, with its valuation governing intersection degree and degeneration asymptotics.","lead":"The paper extends the classical cross-ratio of four points on a line to four planes in projective n-space, proving it equals the height pairing of the corresponding cycles. This provides a unified geometric framework for computing Archimedean and non-Archimedean height pairings in arbitrary dimensions.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"No significant objection identified. The reader's concern about Theorem 2.2 is addressed by the strictness property of morphisms of mixed Hodge structures.","rationale":"The reader identified Theorem 2.2's condensed proof as the weakest point, but this concern does not actually undermine the argument. For extensions of Z(0) by Z(1) — which have only two weight graded pieces — the strictness property of MHS morphisms (Deligne) guarantees that a morphism inducing isomorphisms on graded pieces is an isomorphism of extensions. The paper's argument is therefore complete as stated. The Z(1) check is terse but symmetric to the Z(0) check and uses standard residue maps. I examined the rest of the paper for alternative load-bearing concerns: Proposition 2.1's integral computation is standard Carlson theory; Lemma 4.2's Smith normal form argument is careful with correct telescoping; Theorem C follows cleanly from Proposition 5.1. Independent confirmation by Goncharov (different proof) and Werner (different methods) provides additional support. The verdict of ACCEPT is appropriate.","tokens_in":15184,"tokens_out":4538,"duration_ms":321635,"concrete_test":"To verify the condensed Z(1) check in Theorem 2.2 explicitly: trace the residue maps through the incidence correspondence for the weight-1 graded pieces. Specifically, verify that the composition H^{2k+2}(P^n - B)(k+1) → H^{2k+1}(P^n - B, A)(k+1) [residue/coboundary for H_P] maps under q* and p! to the residue H^2(G - B^≬)(1) → H^1(G - B^≬, [A])(1) [residue for H_G], and that this composition sends the canonical generator (the residue class of the divisor B_1^≬ - B_2^≬, i.e., dlog(b_1^⊥/b_2^⊥)) to itself. This can be checked on the level of differential forms: pull back dlog(b_1^⊥/b_2^⊥) via the incidence correspondence and verify it restricts to the residue form on P^n - B.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest assumption targets Theorem 2.2, worrying that the verification that ϕ: H_P → H_G is an isomorphism of MHS is condensed — only the graded pieces are checked. However, this concern does not land. Both H_P and H_G are extensions of Z(0) by Z(1), meaning they have exactly two non-trivial weight graded pieces (weights 0 and 1). By Deligne's strictness theorem for morphisms of mixed Hodge structures, a morphism of MHS that induces isomorphisms on all weight graded pieces is automatically an isomorphism of MHS (and of extensions). The paper's statement — 'we need only show that the generators of the weight graded pieces map to one another' — is therefore a complete argument, not a gap. The Z(0) piece is checked explicitly via coboundary maps (both q* and p! are shown to be isomorphisms on the relevant H^{2k}(A)(k+1) → H^0([A])), and the Z(1) piece is checked 'analogously' via residue maps. While the Z(1) check is terse, the symmetry of the argument (residue maps play the same role as coboundary maps for the other end of the extension) makes it straightforward. The rest of the paper is also sound: Proposition 2.1's integral computation is standard (Carlson's theory of extensions), Lemma 4.2's Smith normal form argument with Fulton's intersection theory is careful and correct (the telescoping sum properly yields v_R(det μ)), and Theorem C follows cleanly from Proposition 5.1 and multiplicativity of reg_t. Independent confirmation by Goncharov [Gon26] (different proof, inductive) and Werner [Wer01] (different methods) further supports the results.","agreement_with_reader":"disagree"},"referee_report":{"model":"glm-5.2","summary":"This paper generalises the classical cross-ratio of four points on P^1 to a cross-ratio of four linear subspaces in P^n (two k-planes A_1, A_2 and two (n-k-1)-planes B_1, B_2, pairwise disjoint). Three main results are proved: (Theorem A) over C, the cross-ratio equals the augmented height pairing of the cycles A_1-A_2 and B_1-B_2; (Theorem B) over a DVR, the valuation of the cross-ratio equals the intersection degree of the spread-out cycles; (Theorem C) the t-regularised limit of the cross-ratio equals a degenerate cross-ratio on the special fibre. Combining Theorems A and B yields Corollary B', which proves a conjecture of Chen on the asymptotics of Archimedean height pairings for degenerating planes. The proofs combine mixed Hodge theory (Carlson, Hain), an incidence correspondence on the Grassmannian, Smith normal form lattice arguments, and Fulton's intersection theory.","tokens_in":15404,"tokens_out":1969,"duration_ms":156298,"significance":"The paper makes a clean contribution at the interface of Arakelov geometry, mixed Hodge theory, and intersection theory. The generalised cross-ratio is a natural and parameter-free invariant, and the identity with the augmented height pairing (Theorem A) is a satisfying higher-dimensional generalisation of the classical P^1 case. Theorem B provides a concrete, computable formula relating the non-Archimedean local height to an intersection degree, and Corollary B' verifies Chen's conjecture [Che25, Conjecture 1.5] in the setting of degenerating linear subspaces of arbitrary dimension. The degenerate cross-ratio (Definition 1.3) and arithmetic cross-ratio (Definition 5.6) are natural constructions that give a motivic interpretation of limit periods. The independent confirmation by Goncharov [Gon26] via a different method (induction rather than correspondence) strengthens confidence in the results. The Smith normal form argument in Lemma 4.2 is careful and the telescoping computation is correct.","major_comments":[],"minor_comments":[{"comment":"Theorem 2.2, proof: the verification that the Z(1) generator maps correctly is dispatched with 'we argue analogously this time using the residue maps.' While the argument is indeed analogous (and sufficient by Deligne's strictness, since both sides are extensions of Z(0) by Z(1) with only two non-trivial weight pieces), a sentence or two making the residue map check explicit would improve readability for readers less familiar with the correspondence.","section":null},{"comment":"Proposition 2.1: the statement that 'the exponential of this integral is the extension class for H_G' is described as 'a general fact.' A reference to the specific result in Carlson's theory (e.g., [Car80]) that justifies this would be helpful, even if the result is standard.","section":null},{"comment":"Section 3.2: the computation a_2 ∧ b_1 = a_2 ∧ b_2 = (-1)^{k+1} e appears to contain a typo; presumably a_2 ∧ b_j should involve the graph of φ_j, and the two expressions should differ. Please verify this line.","section":null},{"comment":"Remark 4.3 notes that Theorem B also follows from [Wer01, Thm. 3.4]. It would be useful to briefly state what is additionally gained by the present proof (e.g., the explicit Smith normal form decomposition, or the removal of the finite-extension hypothesis), so the reader can assess the relative merits.","section":null},{"comment":"Definition 1.3 and surrounding text: the notation c = (c_{11}, c_{12}, c_{21}, c_{22}) with c_{ij} ∈ hom(det Ŝ_{ij}, det Q̂_{ij}) is introduced somewhat abruptly. A brief remark clarifying that each c_{ij} lives in a potentially different hom-space (since Ŝ_{ij} and Q̂_{ij} depend on the pair (i,j)) would help the reader.","section":null},{"comment":"Corollary C'': the statement involves log|c^⊗_{lim}/c^⊗_{ar}|_σ, but c^⊗_{ar} is defined only up to O^×_κ, so the ratio lives in κ^×/O^×_κ. The absolute value is then well-defined up to log|O^×_κ|_σ, which is consistent with the R/log|O^×_κ|_σ on the right-hand side, but this subtlety should be stated explicitly.","section":null},{"comment":"The reference [Bak26] is the first author's M.S. thesis. While appropriate for context, the paper should be self-contained; the proofs of Theorems A, B, and C here appear to be complete, but the authors should confirm that no step silently relies on the thesis.","section":null},{"comment":"Page 2, line below equation (1.5): 'the logarithms of primes are linearly independent over Q' — this should say 'over Z' or 'Q-linearly independent' for precision, since the statement is about linear combinations with integer coefficients summing to zero.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is solid and the proofs check out. The stress-test concern about Theorem 2.2 does not land: by Deligne's strictness theorem for morphisms of mixed Hodge structures, checking the weight graded pieces is sufficient when both sides are extensions of Z(0) by Z(1). The Z(1) check is terse but correct by symmetry with the Z(0) argument. I see no reason to delay publication beyond minor presentational improvements. The simultaneous appearance with Goncharov's independent proof [Gon26] is handled transparently in Remark 1.2."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful reading and a positive assessment. The referee's recommendation is minor revision, and the report does not raise specific major or minor comments requiring changes to the manuscript. We address the substance of the report below.","responses":[{"response":"We are grateful for the referee's careful and accurate summary of the paper. We confirm that the referee's account of our results, methods, and the relationship to Chen's conjecture [Che25, Conjecture 1.5] is correct. We also note the referee's observation regarding the independent confirmation by Goncharov [Gon26] via a different method (induction rather than correspondence), which we discuss in Remark 1.2. Since no specific issues are raised, no revision is needed on this point.","revision_made":"no","referee_comment":"The referee provides a detailed and accurate summary of the paper's three main results (Theorems A, B, C), the role of Corollary B' in verifying Chen's conjecture, and the methods used (mixed Hodge theory, incidence correspondence, Smith normal form, Fulton's intersection theory). No specific corrections or requests for revision are raised."},{"response":"We appreciate the referee's positive assessment of the paper's significance and the verification of correctness for Lemma 4.2. The recommendation of minor revision is noted. As the referee's report does not identify specific points requiring revision, we will conduct a final careful proofreading of the manuscript to address any typographical or expositional issues before the final version. If the referee has specific minor corrections they would like to flag, we would be happy to incorporate them.","revision_made":"partial","referee_comment":"The referee assesses the paper as making 'a clean contribution at the interface of Arakelov geometry, mixed Hodge theory, and intersection theory,' finds the generalised cross-ratio to be 'a natural and parameter-free invariant,' and evaluates the Smith normal form argument in Lemma 4.2 as 'careful' and the telescoping computation as 'correct.' The recommendation is minor_revision."}],"tokens_in":14773,"tokens_out":470,"duration_ms":61414,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The paper defines a cross-ratio for complementary configurations of planes in P^n and proves three things: it equals the augmented height pairing (Theorem A), its valuation over a DVR equals the intersection degree of spread-out cycles (Theorem B), and a degenerate version captures the limit period of a degenerating family (Theorem C). Corollary B' resolves Chen's conjecture [Che25, Conjecture 1.5] for degenerating planes of any dimension. This is a real result, independently confirmed by Goncharov [Gon26] via a different (inductive) proof and partially anticipated by Werner [Wer01]. The authors are upfront about all of this in Remarks 1.2 and 4.3. The definition itself (Definition 1.1) is natural — it is the obvious generalisation of the classical formula (1.1) — and the connection to height pairings via the incidence correspondence on the Grassmannian is the right way to see it. Proposition 2.1 computes the extension class cleanly using Carlson's theory, and the Smith normal form argument in Lemma 4.2 is careful: the telescoping sum properly yields v_R(det mu), and the use of Fulton's residual formula is correct. The reader flagged the proof of Theorem 2.2 as potentially condensed — only the weight graded pieces are checked, not the full MHS isomorphism. This concern does not land. Both H_P and H_G are extensions of Z(0) by Z(1), so they have exactly two non-trivial weight graded pieces (weights 0 and 1). By Deligne's strictness theorem, a morphism of MHS that induces isomorphisms on all graded pieces is automatically an isomorphism. The paper's statement that 'we need only show that the generators of the weight graded pieces map to one another' is a complete argument, not a gap. The Z(1) check is terse but the symmetry with the Z(0) check makes it straightforward. Theorem C is more definitional than theorem — the authors say so themselves ('the heavy lifting is done by the definitions') — but the construction of the limit perturbation c_lim in Proposition 5.1 is clean, and the arithmetic cross-ratio and its norm formula (Lemma 5.7) are sensible. The restriction to class number 1 for the arithmetic version is a real limitation but the authors note the fractional ideal generalisation in Remark 5.4. This paper is for arithmetic geometers and Hodge theorists working on height pairings, regulators, or degeneration questions. It is short, self-contained, and the proofs check out. It deserves a serious referee, primarily to verify the Fulton intersection theory details in Section 4 and the determinantal identities in Section 5, which are the parts where a careful reader would want to see every step.","headline":"Clean, correct paper giving a generalised cross-ratio equal to the augmented height pairing, with a valuation formula and degenerate extension. Proofs are sound; the one flagged concern about Theorem 2.2 does not land.","tokens_in":16222,"tokens_out":680,"would_cite":true,"duration_ms":106672,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14G40","14C17","32G20","14C30","14N20","14C25"],"pacs":[],"model":"glm-5.2","headline":"Generalized cross-ratio unifies height pairings and intersection degrees","keywords":["cross-ratio","height pairing","mixed Hodge structure","intersection theory","Grassmannian","degeneration","local heights","arithmetic geometry"],"falsifier":"If the incidence-correspondence map between H_P and H_G fails to be an isomorphism of mixed Hodge structures (not just at the graded-piece level), Theorem A would not follow, and the entire chain of identifications linking the cross-ratio to height pairings, valuations, and intersection degrees would break.","tokens_in":15254,"feed_emoji":"✦","tokens_out":1532,"duration_ms":137075,"temperature":0.7,"pith_summary":"The paper extends the classical cross-ratio of four points on a line to a cross-ratio of four linear subspaces in projective n-space, where two subspaces are k-dimensional and two are (n-k-1)-dimensional, with each pair from opposite groups being disjoint. The authors prove three things about this generalized cross-ratio. First, over the complex numbers, it equals the augmented height pairing of the corresponding cycles (Theorem A), a direct higher-dimensional analogue of the classical fact that the cross-ratio of four points equals their height pairing. Second, over a discretely valued field, the valuation of the cross-ratio equals the intersection degree of the cycles when spread out over the valuation ring (Theorem B). Third, when the four subspaces are allowed to intersect, a degenerate cross-ratio can be defined using infinitesimal perturbation data, and the limiting value of the height pairing in a degenerating holomorphic family equals this degenerate cross-ratio on the central fibre (Theorem C). Together, these results show that the asymptotic behavior of Archimedean height pairings in degenerating families is governed by a purely algebro-geometric intersection degree, and that the limiting constant term admits a motivic interpretation as a degenerate cross-ratio.","feed_headline":"Generalized cross-ratio governs height pairings in all dimensions","feed_subtitle":"Four planes in projective space get a cross-ratio that unifies Archimedean heights, p-adic valuations, and intersection degrees into one公式.","key_machinery":"The cross-ratio is defined as an alternating product of pairings between generators of exterior powers of subspaces and their duals. The proof of Theorem A uses an incidence correspondence between the Grassmannian G(k,n) and projective space P^n to construct an isomorphism of mixed Hodge structures. The proof of Theorem B uses Smith normal form and Fulton's intersection theory for degeneracy loci. The degenerate cross-ratio relies on a determinantal identity (det V ≅ det A ⊗ det B ⊗ hom(det S, det Q)) and a regularisation map that extracts the leading coefficient of a valued element.","core_discovery":"The generalized cross-ratio of four complementary linear subspaces in projective n-space serves as a single object that simultaneously encodes the Archimedean height pairing (over C), the non-Archimedean local height (via valuation), and the intersection degree (over a DVR). The key mechanism is that the cross-ratio, defined as an alternating product of determinantal pairings between subspaces and their annihilators, reduces to the classical cross-ratio when n=1, and the authors prove it coincides with the extension class of a mixed Hodge structure via an incidence correspondence between the Grassmannian and projective space. This identification, combined with the valuation-intersection dier","pith_inferences":["The cross-ratio's role as a universal invariant bridging Archimedean and non-Archimedean heights suggests that other classical projective invariants might admit similar height-theoretic interpretations in higher dimensions, particularly for configurations involving more than four subspaces.","The regularisation map reg_t, which extracts the leading coefficient of a valued element, is analogous to regularisation procedures for period integrals; this connection could potentially unify algebraic and analytic approaches to limit periods.","The dependence of the limit perturbation c_lim only on t mod m^2 suggests that the leading asymptotics of height pairings are determined by first-order deformation data, which could simplify numerical computations in practice.","The fact that Theorem B requires no properness assumption, combined with Corollary B', hints that Chen's conjecture might hold for arbitrary (not just proper) degenerations of algebraically trivial cycles beyond the linear case."],"forward_implications":["Corollary B' confirms Chen's conjecture (that divergence of Archimedean height pairings in degenerating families is governed by intersection degree) for the case of degenerating linear subspaces of arbitrary dimension, without requiring properness assumptions.","The degenerate cross-ratio provides a framework for computing limit periods of mixed Hodge structures in arbitrary-dimensional degenerations with arbitrarily bad singularities, offering a testbed for more general limit-period conjectures.","The arithmetic cross-ratio and its norm expression in terms of lattice indices connect Archimedean limit periods to non-Archimedean arithmetic data, suggesting a path toward explicit height computations on moduli spaces of linear subspace configurations.","The S4 symmetry of the cross-ratio in the middle-dimensional case (n=2k+1) may yield new constraints on height pairings and their behavior under duality, analogous to classical cross-ratio symmetries."],"fun_headline_variants":["Cross-ratio generalized to four planes unifies local heights","Four-plane cross-ratio encodes height pairings and intersection degree","Higher-dimensional cross-ratio governs Archimedean and p-adic heights","Generalized cross-ratio links local heights to intersection degrees","Plane cross-ratios determine intersection degrees and local heights"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The proof of Theorem A depends on showing that a map between two mixed Hodge structures, constructed via an incidence correspondence between the Grassmannian and projective space, is an isomorphism. The authors verify this at the level of the weight-graded pieces (the top and bottom layers of the structure) but the full verification that the map respects the complete mixed Hodge structure is condensed; if this map fails to be a genuine isomorphism of mixed Hodge structures, a","fun_headline_variants_meta":{"raw":{"variants":["Cross-ratio generalized to four planes unifies local heights","Four-plane cross-ratio encodes height pairings and intersection degree","Higher-dimensional cross-ratio governs Archimedean and p-adic heights","Generalized cross-ratio links local heights to intersection degrees","Plane cross-ratios determine intersection degrees and local heights"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":1198,"prompt_tokens":555,"completion_tokens":643,"prompt_tokens_details":null},"tokens_in":555,"tokens_out":643,"duration_ms":44821,"temperature":1.0,"reasoning_tokens":588,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T04:27:12.709723+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If the incidence-correspondence map between H_P and H_G fails to be an isomorphism of mixed Hodge structures (not just at the graded-piece level), Theorem A would not follow, and the entire chain of identifications linking the cross-ratio to height pairings, valuations, and intersection degrees would break.","supporting_citations":[],"review_version":1}