{"id":"d0f3cdcf-3b75-46bf-a30b-54165d8381cc","arxiv_id":"2505.05925","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For infinite locally finite cellular decompositions in spherical background geometry, long-time existence and, under an additional condition, convergence of the prescribed-total-geodesic-curvature Ricci flow are established.","lead":"This paper proves that a combinatorial Ricci flow for circle patterns on infinite surfaces exists for all time in spherical background geometry, and converges to prescribed curvatures under an extra initial condition. It is the first infinite version of this flow in the spherical setting, where the usual convexity arguments fail.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.2 applies Lemma 2.2 with a uniform row-sum bound (3.19) that is not established by the preceding estimates; the convergence proof is incomplete as written, though the gap is likely repairable by a finite maximum principle.","rationale":"The reader's weakest-assumption analysis identifies the same obstruction I find. The existence theorem (1.2) is supported by a reasonable finite-exhaustion and Arzelà-Ascoli argument; per-vertex C^2 bounds suffice because the diagonalization is done vertex by vertex. The convergence theorem (1.3) is the delicate part. Its monotonicity step needs f^n≥0 for the finite flows. The paper proves this by applying Lemma 2.2, and Lemma 2.2 cannot be applied unless (3.19) holds uniformly in i and n. The cited estimates (3.10)-(3.11) are only per-vertex; local finiteness and (S1)-(S3) impose no uniform degree or angle lower-bound, so (3.19) is genuinely unsupported. However, the same conclusion can plausibly be obtained by a finite maximum principle on V[n], using that f^n is supported in V[n], ω_ij≥0, and g_i≤0 from (2.9). Thus the correct assessment is not that Theorem 1.3 is false, but that the written proof has an identifiable, likely repairable gap. This matches the reader's CONDITIONAL verdict, so I recommend no change.","tokens_in":10814,"tokens_out":25093,"duration_ms":251902,"concrete_test":"Test the gap by proving f^n(t)≥0 directly on the finite set V[n]: at the first time t0 at which min_{i∈V[n]} f_i^n(t)=0, use (3.18) to compute the derivative of the minimum; ω_ij≥0 and g_i≤0 (which follows from (2.9)) should force the derivative to be nonnegative, so f^n cannot cross below 0. No uniform row-sum bound is used. If this argument succeeds, Theorem 1.3 survives with the finite maximum principle replacing Lemma 2.2; if it encounters a term requiring a bound uniform in i and n, then the missing estimate (3.19) is essential.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1.3 rests on Theorem 3.2, whose key step is the claim f^n(t)≥0 for the finite approximations. This is obtained by applying Lemma 2.2 to (3.18). Lemma 2.2 explicitly requires a uniform constant C with ∑_{j∼i}ω_ij(t)<C for every vertex i and every t∈[0,τ]. The paper asserts at (3.19) that this follows from (3.10)-(3.11). Those estimates, however, bound derivatives of T_i^n only for each fixed vertex i, with constants depending on i, on deg(i), and on the incident angles Θ(e); they provide no bound that is uniform over i∈V[n] as n grows. Local finiteness and conditions (S1)-(S3) do not imply a uniform degree bound, and the formula (2.8) for ω_ij in terms of sinΘ(e) shows that any uniform estimate would need a separate argument. Thus (3.19) is not proved, and the maximum-principle conclusion f^n≥0, hence monotonicity of u(t) and the convergence claim of Theorem 1.3, is not justified as written. The gap is likely patchable: f^n has finite support in V[n], and by (2.9) the coefficient g_i=-(∂f_i/∂u_i+∑_{j∼i,v_j∈V[n]}∂f_i/∂u_j) is ≤0, so a finite maximum principle on V[n] would yield f^n≥0 without any uniform row-sum bound. But this replacement is not in the manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the combinatorial Ricci flow with prescribed total geodesic curvatures for circle patterns in spherical background geometry, in the setting of infinite locally finite cellular decompositions. The main results are Theorem 1.2, asserting long-time existence of the flow for arbitrary initial data under conditions (S1) and (S2), and Theorem 1.3, asserting convergence to the prescribed total geodesic curvature under the additional initial monotonicity condition (S3). The proof approximates the infinite graph by an exhausting sequence of finite subcomplexes, uses the finite-dimensional ODE theory to obtain solutions on each approximating complex, and then extracts a diagonal subsequence by Arzelà-Ascoli estimates to obtain a global solution. For convergence, the paper applies a maximum principle for infinite graphs to the difference f^n = T^n - T_hat on each finite approximation, aiming to show that f^n remains nonnegative and hence that the flow is monotone.","tokens_in":11168,"tokens_out":11911,"duration_ms":121833,"significance":"If the results are correct, this is the first treatment of an infinite combinatorial curvature flow in spherical background geometry, and the existence theorem extends the finite-dimensional theory of Ge-Hua-Zhou in a natural way. The proof does not assume its conclusion: the limit solution is constructed from finite approximations, and the maximum principle is proved independently as a lemma. The potential-function framework from Nie and Ge-Hua-Zhou is used only as background, and conditions (S1)-(S3) are hypotheses rather than fitted outputs. However, the convergence proof currently relies on a uniform bound that is not established, so the central theorem is not yet justified as written.","major_comments":[{"comment":"The uniform row-sum bound (3.19) is asserted but not proved. The preceding estimates (3.10)-(3.11) bound derivatives of T_i^n only for each fixed vertex i, with constants depending on i, on the degree of i, and on the incident angles; they provide no bound uniform over all vertices as n grows. Nothing in local finiteness, (S1), (S2), or (S3) prevents the vertex degrees from being unbounded or sin(Theta(e)) from approaching zero. Since the weights omega_ij in (3.16) are given by -partial f_i^n/partial u_j, and formula (2.8) involves a factor 1/sin(Theta(e)), the row sums can be unbounded. Therefore Lemma 2.2 cannot be applied as written, and the conclusion f^n(t) >= 0, which is the basis for monotonicity of u(t) and the convergence claim of Theorem 1.3, is not justified.","section":"Section 3, Theorem 3.2, equations (3.18)-(3.19)"},{"comment":"The same gap affects the claimed uniform bound on g. The statement 'by (3.11), we know |g| <= C0 for an uniform constant C0' is not justified by (3.10)-(3.11), which are vertex-dependent; indeed g_i is a sum over the neighbors of i, so its magnitude can grow with the degree. A related but separate point is that the manuscript does not use the sign information available from (2.9) to show g_i <= 0. The argument could likely be repaired by applying a finite maximum principle on V[n] to the finitely supported function f^n, using g_i <= 0 and omega_ij >= 0, but such an argument is not included in the paper.","section":"Section 3, Theorem 3.2, paragraph after (3.18)"}],"minor_comments":[{"comment":"The citation 'see [21]' for profound results in the infinite setting is apparently incorrect, since reference [21] is Hamilton's Ricci flow paper and not a circle-pattern reference; the intended citation may be [13] or [22].","section":"Section 1.1, paragraph on infinite settings"},{"comment":"The statement of the maximum principle for infinite graphs does not explicitly assume omega_ij(t) >= 0, but the proof uses this nonnegativity when it asserts that Delta_G f_delta <= 0 at a maximum; the application in the paper does have nonnegative weights, but the lemma as stated is incomplete.","section":"Lemma 2.2"},{"comment":"The phrasing 'where constant C only depends on tau, i, j where j ~ i and e in E(v_i)' is confusing; it should say that the constant depends on the fixed vertex i, its neighbor j, and the incident edges, and it is not uniform in i.","section":"Equation (3.10)"},{"comment":"The diagonal-order argument is terse; the definition of the predecessor in the path through N^2 could be clarified, and a short explanation of why the diagonal sequence lies eventually in each chosen subsequence would improve readability.","section":"Theorem 3.1, diagonal subsequence construction"},{"comment":"There is a typo: 'the functions(r) is given by s(r)' should read 'the function s(r) is given by s(r)'.","section":"Section 1.1, first paragraph"}],"recommendation":"major_revision","confidential_remarks":"The gap identified in the convergence proof is genuine but appears reparable: since f^n has finite support in V[n], a standard finite maximum principle with g_i <= 0 should give f^n >= 0 without any uniform row-sum bound. If the authors adopt this repair, Theorems 1.2 and 1.3 are likely correct. The paper's novelty claim of being the first infinite combinatorial curvature flow in spherical background geometry appears reasonable in light of the cited literature, including the concurrent preprint [13], which concerns Euclidean and hyperbolic settings."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the first infinite combinatorial Ricci flow in spherical background geometry. That claim checks out: prior infinite circle pattern flow work of He-Schramm, He, and Ge-Hua-Zhou lives in Euclidean or hyperbolic settings, and spherical background genuinely breaks the convexity that made those arguments work. The paper deserves credit for identifying and attacking that gap.\n\nThe existence theorem, Theorem 1.2, is well supported. The finite exhaustion with bounded C^2 estimates and a diagonal Arzela-Ascoli subsequence is standard but correctly executed. The bounds for fixed vertices come from Gauss-Bonnet and are local, so the limit passes through cleanly. I do not see a real problem here.\n\nThe convergence theorem, Theorem 1.3, has a soft spot in the proof of monotonicity. The argument applies Lemma 2.2 on the infinite graph to each f^(n), and Lemma 2.2 requires a uniform row-sum bound sum_{j~i} omega_ij(t) < C for all vertices. The proof asserts at (3.19) that this follows from (3.10)-(3.11), but those estimates carry constants depending on i, on deg(i), and on 1/sin(Theta(e)). Local finiteness and (S1)-(S3) do not give uniformity. So the maximum principle step is not justified as written. That is a real gap.\n\nIt is, however, likely patchable. For each fixed n, the function f^(n) has finite support in V[n] and the coefficients are bounded on that finite graph. A finite maximum principle, or even the cooperative structure of the linear system with nonnegative off-diagonal weights, should preserve the sign of f^(n)(0) without any uniform-in-i row-sum bound. The paper does not contain that replacement, but the fix is not deep.\n\nI also note that (S2) is stated but never used in the convergence proof; that is harmless but could be flagged in revision. The citation pattern is honest and priorities are clearly stated. The paper is not overselling itself; the open problems are natural.\n\nWho is this for? Mathematicians working in discrete conformal geometry and circle patterns, especially anyone interested in noncompact surfaces and parabolic methods. It is not a breakthrough that reshapes the field, but it is a meaningful step into territory that was previously inaccessible.\n\nRecommendation: send it to peer review. A good referee should ask for the missing uniform bound or for an explicit finite maximum principle. With that repaired, the paper is publishable.","headline":"First infinite spherical combinatorial Ricci flow with a solid existence proof; the convergence proof has a real but likely repairable gap in the maximum principle application.","tokens_in":11636,"tokens_out":4962,"would_cite":true,"duration_ms":52641,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["52C26","53C44"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the combinatorial Ricci flow in spherical background geometry exists for all time on infinite locally finite cellular decompositions, and converges to prescribed total geodesic curvatures under a comparison…","keywords":["combinatorial Ricci flow","circle patterns","spherical background geometry","total geodesic curvature","infinite cellular decomposition","maximum principle","global existence","convergence"],"falsifier":"Construct an infinite locally finite decomposition satisfying (S1)-(S3) with vertex degrees growing without bound and edge angles $\\Theta(e)\\to0$. For a fixed $\\tau>0$, compute the conductances $\\omega_{ij}(t)$ from the finite approximations and check whether $\\sup_{t\\in[0,\\tau]}\\sup_i\\sum_{j\\sim i}\\omega_{ij}(t)$ is finite. If it is infinite, the maximum-principle step in Theorem 3.2 cannot be applied as written; if the flow still converges, the theorem is true but needs a different proof, and if it fails to converge, Theorem 1.3 is false.","tokens_in":10613,"feed_emoji":"⭕","tokens_out":9722,"duration_ms":103719,"temperature":0.7,"pith_summary":"This paper establishes global existence and, under an extra comparison condition, convergence for the combinatorial Ricci flow with prescribed total geodesic curvatures on infinite surfaces in spherical background geometry. The vertex set is infinite, so the usual Picard-Lindelof existence theory does not apply; the authors instead approximate the infinite cellular decomposition by an increasing sequence of finite subcomplexes, solve the flow on each, and extract a limit solution by a diagonal subsequence argument. The convergence result follows from a maximum principle on infinite graphs applied to the difference between current and target total geodesic curvature. Because prior infinite circle-pattern results were confined to Euclidean and hyperbolic geometry, this is the first such flow theorem in spherical background geometry.","feed_headline":"Infinite spherical circle-pattern flow exists and converges","feed_subtitle":"For infinite cell decompositions, a curvature flow starting above its target settles at the prescribed geodesic curvature.","key_machinery":"The load-bearing mechanism is the sign and symmetry structure of the edge-wise total geodesic curvature $T(e,v)$. For an edge $e=\\{v_i,v_j\\}$, Lemma 2.1 gives $\\partial T(e,v_i)/\\partial u_j = \\partial T(e,v_j)/\\partial u_i$, with mixed derivatives negative and the sum derivative positive. This turns the flow into a gradient-like system for the convex potential $E(u)=\\sum_e E_e(u_i,u_j)-\\sum_v\\hat{T}_v u_v$, and, when $T\\ge\\hat{T}$, makes the difference $f_i=T_i-\\hat{T}_i$ satisfy a parabolic inequality $df_i/dt=\\Delta_\\omega f_i + g_i f_i$ with nonnegative conductances $\\omega_{ij}=-\\partial T_i/\\partial u_j$. The infinite maximum principle (Lemma 2.2) then forces $f_i\\ge0$ for all time, giving monotonicity of $u(t)$; the lower bound on $u_i$ comes from $\\hat{T}_i>0$ and $T_i=\\alpha_i\\cos r_i\\le \\pi\\deg(v_i)\\cos r_i$. The finite-exhaustion and diagonal-subsequence construction is what bypasses the failure of classical ODE theory on infinite vertex sets.","core_discovery":"The central claim is Theorem 1.2 and Theorem 1.3. For an infinite locally finite cellular decomposition $D=(V,E,F)$ with intersection angles $\\Theta(e)\\in(0,\\pi/2]$, and target total geodesic curvatures $\\hat{T}_v$ satisfying (S1) $\\hat{T}_v>0$ for every vertex and (S2) $\\sum_{v\\in U}\\hat{T}_v<2\\sum_{e\\in E(U)}\\Theta(e)$ for every finite $U\\subset V$, the flow $$\\frac{du_i}{dt}=-(T_i-\\hat{T}_i),\\qquad u_i=\\ln\\cot r_i,$$ has a solution $u(t)$ for all $t\\ge 0$. If additionally the initial data satisfy (S3) $T(r(0))\\ge \\hat{T}$, then this solution converges as $t\\to\\infty$, and $\\lim_{t\\to\\infty}T(r(t))=\\hat{T}$. The proof constructs the infinite-time solution by exhausting $D$ with finite complexes $D[n]$, solving the finite flow on each, and taking a diagonal subsequence; the convergence proof shows the solution is monotone via a maximum principle and uses the fact that $T_i=\\alpha_i\\cos r_i$ forces $u_i$ to have a lower bound.","pith_inferences":["A testable extension is to allow intersection angles $\\Theta(e)\\in(0,\\pi)$, the range raised as an open problem; the proof should carry through if the sign conditions in Lemma 2.1 remain valid above $\\pi/2$.","The finite-exhaustion scheme is local, so it likely applies to other infinite cellular decompositions on noncompact or nonorientable surfaces, not only disk triangulations.","If the uniform row-sum bound required by the maximum principle is not automatic from the stated hypotheses, a localized maximum principle with cut-off functions may still prove convergence while needing only local bounds.","For numerical construction of infinite spherical circle patterns, Theorem 1.3 offers a stopping criterion: start with $T(0)\\ge\\hat{T}$, and once $T(r(t))-\\hat{T}$ is uniformly small, the remaining drift toward the target is controlled by that difference."],"forward_implications":["Given (S1) and (S2), the flow can be started from any initial radii in $(0,\\pi/2)$ and will never leave that range, so no finite-time blow-up or boundary collision occurs.","Under (S3), the solution is monotone in the variable $u_i=\\ln\\cot r_i$, and the limiting radii give a spherical circle pattern with prescribed total geodesic curvatures.","The theorem provides a parabolic construction of infinite circle patterns in spherical geometry, extending the finite-cell-decomposition result to noncompact surfaces.","The diagonal-subsequence argument shows the infinite flow is well-defined and independent of the choice of finite exhaustion, since limits on common time intervals agree."],"supporting_citations":[{"why":"introduced the combinatorial Ricci flow and its finite global-existence and convergence theory, the framework this paper extends","marker":"[2]"},{"why":"defined the spherical-background prescribed-total-curvature flow (2.2) and proved the finite analogue with conditions (S1)-(S2)","marker":"[12]"},{"why":"supplied the convex potential function and the edge derivative identities (Lemma 2.1) that turn the flow into a gradient-like system","marker":"[27]"},{"why":"showed how combinatorial Ricci flow can be run on infinite disk triangulations by finite exhaustion, the technique adapted here","marker":"[13]"},{"why":"provided the infinite-disk-pattern rigidity background that motivates extending circle-pattern flows to infinite surfaces","marker":"[23]"}],"fun_headline_variants":["Infinite spherical Ricci flow exists and converges","First infinite spherical Ricci flow study","Spherical Ricci flow solves infinite cell decompositions","Infinite spherical Ricci flow exists and settles","Beyond finite: spherical Ricci flow exists and converges"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The convergence argument presumes that, at every vertex and every time, the total influence of a vertex's neighbors through the discrete Laplacian stays below one universal constant, and the stated hypotheses do not by themselves obviously guarantee that.","fun_headline_variants_meta":{"raw":{"variants":["Infinite spherical Ricci flow exists and converges","First infinite spherical Ricci flow study","Spherical Ricci flow solves infinite cell decompositions","Infinite spherical Ricci flow exists and settles","Beyond finite: spherical Ricci flow exists and converges"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00152,"raw_usage":{"total_tokens":6070,"prompt_tokens":909,"completion_tokens":5161,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":525,"completion_tokens_details":{"reasoning_tokens":5095}},"tokens_in":525,"tokens_out":5161,"duration_ms":42735,"temperature":1.0,"reasoning_tokens":5095,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:54:35.115171+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct an infinite locally finite decomposition satisfying (S1)-(S3) with vertex degrees growing without bound and edge angles $\\Theta(e)\\to0$. For a fixed $\\tau>0$, compute the conductances $\\omega_{ij}(t)$ from the finite approximations and check whether $\\sup_{t\\in[0,\\tau]}\\sup_i\\sum_{j\\sim i}\\omega_{ij}(t)$ is finite. If it is infinite, the maximum-principle step in Theorem 3.2 cannot be applied as written; if the flow still converges, the theorem is true but needs a different proof, and if it fails to converge, Theorem 1.3 is false.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"introduced the combinatorial Ricci flow and its finite global-existence and convergence theory, the framework this paper extends"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"defined the spherical-background prescribed-total-curvature flow (2.2) and proved the finite analogue with conditions (S1)-(S2)"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplied the convex potential function and the edge derivative identities (Lemma 2.1) that turn the flow into a gradient-like system"}],"review_version":1}