{"id":"1025e13b-cce8-431c-9576-1ed0300f503b","arxiv_id":"2608.13235","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Under E[M_-^{(d+δ)/2}]<∞, the Laguerre tessellation generated by a stationary marked point process is a.s. well-defined and preserves the generator's ergodicity, mixing, and alpha-mixing.","lead":"This paper proves that ergodicity, mixing, and alpha-mixing of a stationary marked point process are inherited by the random Laguerre tessellation it generates, under a mild moment condition on the negative marks. The result gives the first general mixing theory for weighted random tessellations beyond the Poisson case.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 1(c) relies on Lemma 12, whose second case contains a scaling error: from z∈rB_j and d(y,z)>β_l r it does not follow that y∉β_l r B_j. The α-mixing preservation proof is therefore incomplete as written.","rationale":"The reader's weakest assumption concerns the 'nearly optimal' moment condition and labels it a minor overclaim. My reading identifies a different, more serious issue inside the proof of the alpha-mixing preservation theorem. The central claim of the paper is Theorem 1, and part (c) is a substantial component of that claim. Its proof depends on Lemma 12 through Lemma 14. Lemma 12's second case contains an invalid containment step: the inequality d(y,z)>β_l r combined with z∈rB_j does not imply y∉β_l r B_j under the paper's own notation rB_j:=B(rx,ra), because β_l r B_j is centered at β_l r x_j rather than at z. Even with the natural correction B(rx_j,β_l r), the stated conclusion is false. This is not an issue of disagreement with existing consensus or a matter of an unstated stronger assumption; it is an internal gap in the written proof. I am not claiming the theorem is false; the arguments in Lemmas 10–11 and the overall approximation strategy suggest the result is likely correct after a repair. But because the submitted proof of Theorem 1(c) rests on an unproven and apparently incorrect lemma, the appropriate verdict is conditional acceptance pending correction and verification of Lemma 12 and the subsequent bound in Lemma 14.","tokens_in":20389,"tokens_out":26853,"duration_ms":270435,"concrete_test":"Independently re-derive Lemma 12, Case 2, at the set-theoretic level. Fix l, r, the cover B_1,…,B_θ_l, take z∈rB_j, and choose y inside β_l r B_j but on the side of r x_j opposite z (for instance y=β_l r x_j when z≈r x_j). Check whether d(y,z)>β_l r can hold while y remains in the claimed excluded set; if it can, the inference is invalid. Then determine the correct excluded ball, e.g. B(rx_j,(β_l−1)r), and verify whether Lemma 14 still yields the required empty-space estimate after replacing every occurrence of β_l r B_j accordingly.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 4.3 derives Theorem 1(c) through Lemmas 10–14, and Lemma 14 bounds P(η∈E^3_{r,K3}) using Lemma 12. In Lemma 12, the notation rB_j:=B(rx,ra) is fixed, and β_l r B_j means B(β_l r x_j, β_l r a) (here a=1). In Case 2 the proof has established only that any generator (y,t) with t≤r^2 satisfies d(y,z)>β_l r, where z is a point of the cell with z∈rB_j. It then asserts that y∉β_l r B_j. This does not follow: β_l r B_j is centered at β_l r x_j, not at z, so the two balls are spatially separated and B(β_l r x_j, β_l r) is not contained in B(z,β_l r). Even under the alternative reading B(rx_j,β_l r), the implication fails: y may lie on the side of rx_j opposite z, giving d(y,z)>β_l r while y is still inside B(rx_j,β_l r). The valid conclusion would instead be a bound on B(rx_j,(β_l−1)r), since d(y,z)≤d(y,rx_j)+d(rx_j,z). As written, Lemma 12 is not justified, so the estimate in Lemma 14 that controls P(η∈E^3_{r,K3}) and hence the final alpha-mixing inequality is not established. The central theorem may still be true and the gap may be repairable by correcting the excluded ball, but the proof of Theorem 1(c) as submitted has a load-bearing gap.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies random Laguerre tessellations generated by a stationary marked point process on R^d x R, with possibly unbounded weights. Its main result, Theorem 1, states that under the moment condition E[M_-^{(d+δ)/2}]<∞ for some δ>0, ergodicity, mixing, and α-mixing of the generating marked point process are each preserved by the Laguerre tessellation map, and that the generating process is then almost surely an admissible generator. Section 3 introduces tempered configurations and proves measurability of the Laguerre mapping (Theorem 4). Section 4 proves the main theorem, with the α-mixing preservation argument based on geometric lemmas controlling large cells via empty regions. Section 5 gives the Poisson case, where the optimal moment condition E[M_-^{d/2}]<∞ is obtained, and discusses several non-Poissonian examples.","tokens_in":20768,"tokens_out":14968,"duration_ms":158033,"significance":"If the proof gaps described below are repaired, the paper would be a valuable contribution: it extends mixing preservation results beyond Poisson and Gibbs-type generators to general stationary marked point processes, provides the first formal measurability proof for the Laguerre diagram map, reproduces the needed argument from the authors' earlier note [26] in full, and gives a clean optimal moment condition in the Poisson case (Lemma 17). The paper is largely self-contained and its main claims are concrete and falsifiable. However, two load-bearing proof steps currently fail as written, so the manuscript is not yet ready for acceptance.","major_comments":[{"comment":"The proof of Lemma 12 contains a scaling error in the second case. From z∈rB_j and ||z−y||>√(C_l^2−1)r one cannot conclude that y∉β_l r B_j, because β_l r B_j = B(β_l r x_j, β_l r) is not centered at z; it may contain points whose distance from z is much larger than √(C_l^2−1)r. The valid triangle-inequality conclusion is instead y∉B(r x_j, β_l r), i.e. y∉r(β_l B_j), which is a different ball. Since Lemma 14 uses exactly the stated conclusion to bound P(η∈E^3_{r,K3}) by a sum of empty-ball probabilities, the proof of Theorem 1(c) is incomplete as submitted. The argument appears repairable by replacing β_l r B_j with r(β_l B_j) in the statement and proof of Lemma 12 and adapting Lemma 14 accordingly, but this must be carried out explicitly.","section":"§4.3, Lemma 12"},{"comment":"The measurability proof discards the sentinel values B(0,1) and B(0,2) in the definition of κ, but these closed balls can themselves be genuine Laguerre cells with non-empty interior. For example, a generator at the origin together with suitably weighted generators in directions dense on the sphere can produce an origin cell that is exactly a ball. For such configurations the equality L(φ)=∪_n L_n^1(φ) is false, because the genuine cell is removed. Therefore the proof of Theorem 4, which is invoked in the proofs of Theorem 1(a) and (b), is not valid as written. The issue is local and can be fixed, for instance by choosing dummy values that cannot coincide with any non-empty-interior Laguerre cell, or by carrying an indicator of whether the value is a dummy.","section":"§4.1, proof of Theorem 4"}],"minor_comments":[{"comment":"The phrase \"nearly optimal\" is overstated. The 'if' direction of Lemma 17 uses only Campbell's formula and does not require the Poisson assumption, so E[M_-^{d/2}]<∞ already suffices for well-definedness for arbitrary stationary marked point processes; the stronger moment (2) in Theorem 1 is needed for the tempered-configuration estimates used in the α-mixing part, not for well-definedness itself.","section":"Abstract and §1"},{"comment":"The sentence introducing (10) is a side remark and could be integrated more cleanly with the surrounding proof, since the proof of part (b) has already been completed at that point.","section":"§4.2, after proof of Theorem 1(b)"},{"comment":"In the 'only if' direction, the notation ρ(0,x) is used before being explicitly defined in the Laguerre formalism section; a brief reminder of the definition of power distance would improve readability.","section":"§5.1, Lemma 17"}],"recommendation":"major_revision","confidential_remarks":"Both main issues are localized and I expect they can be repaired within the scope of the paper: the geometric lemma needs a corrected excluded ball, and the measurability proof needs sentinel values that cannot coincide with genuine cells. I do not see evidence of a fundamental unsoundness in the overall strategy."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a genuine step forward for random Laguerre tessellations: it gives the first general preservation results for ergodicity, mixing, and α-mixing under the Laguerre map, and it proves a measurability theorem for the Laguerre mapping that earlier work had silently assumed. The tempered-configuration machinery for unbounded weights is substantial and well motivated, and the vacancy lemmas in Section 4.3 are clever. The moment condition (2) is cleanly stated, and the paper is honest that the δ>0 is a technical price beyond the sharp d/2 moment in the Poisson case.\n\nThe stress-test note about Lemma 12 is on target. In the second case, the proof establishes that any generator with t≤r^2 lies outside the ball centered at r x_j of radius β_l r, but the lemma states the exclusion for the ball centered at β_l r x_j with radius β_l r. Those are different balls; the stated version is false, while the proof's version is correct. Lemma 14 then uses the stated version, so the written proof of Theorem 1(c) does have a genuine gap at that point. The good news is that the repair is immediate: replace β_l r B_j by B(r x_j, β_l r) in Lemma 12, the proof, and Lemma 14. Lemma 13 still applies because the radius is a fixed multiple of r, so the asymptotic argument survives unchanged. The central theorem is not in danger, but the manuscript as submitted is not fully correct in this section.\n\nA smaller issue: the abstract calls the moment condition “nearly optimal,” but the paper itself shows in Lemma 17 that d/2 suffices for well-definedness of a stationary marked point process, so the (d+δ)/2 condition is not close to necessary in general. That is a minor overstatement, not a flaw in the mathematics.\n\nThis paper is for stochastic geometers and point process theorists, and it definitely deserves a serious referee. I would not desk-reject it. My recommendation: send to referee, and specifically ask the referee to check the ball-centering in Lemma 12 and Lemma 14. With that removed, the paper is a solid contribution.","headline":"Solid paper on mixing preservation for Laguerre tessellations, with one real but repairable notation gap in the α-mixing proof.","tokens_in":21270,"tokens_out":7185,"would_cite":true,"duration_ms":46883,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60D05","37A25","60G55"],"pacs":[],"model":"deepseek-v4-flash","headline":"Random Laguerre tessellations inherit ergodicity, mixing, and α-mixing from the stationary marked point processes that generate them, under a nearly optimal moment bound on negative weights.","keywords":["random Laguerre tessellation","ergodicity","mixing","alpha-mixing","marked point process","tempered configurations","power distance","tessellation"],"falsifier":"Check the $\\alpha$-mixing transfer at the endpoint: take an $\\alpha$-mixing marked point process whose typical mark has density proportional to $|m|^{-1-d/2}/\\log^2 |m|$ for $m\\le -e$. This mark distribution satisfies $E[M_-^{d/2}]<\\infty$ but $E[M_-^{(d+\\delta)/2}]=\\infty$ for every $\\delta>0$. If simulated Laguerre tessellations from such generators still have $\\alpha$-mixing coefficients tending to zero, the strict $\\delta>0$ in Theorem 1 is an artifact of the proof; if not, it is essential.","tokens_in":20227,"feed_emoji":"📐","tokens_out":7939,"duration_ms":78275,"temperature":0.7,"pith_summary":"Random Laguerre tessellations—weighted generalizations of Voronoi diagrams in which each cell is defined by power distance from a nucleus with a weight—are shown to be well defined and to inherit the asymptotic-independence structure of the point process that generates them. The sufficient condition is a moment bound on the negative part of the weights: very large negative weights must be sufficiently rare. Under this condition, an ergodic marked point process produces an ergodic Laguerre tessellation, a mixing process produces a mixing tessellation, and an α-mixing process produces an α-mixing tessellation. This matters for applications in materials science and biology, where Laguerre tessellations model foams, polycrystals, and cellular structures and where one wants spatial averages to converge and far-apart regions to behave independently.","feed_headline":"Random Laguerre tessellations inherit mixing from their generators","feed_subtitle":"A moment condition on negative weights carries ergodicity, mixing, and α-mixing from generator to tessellation.","key_machinery":"The machinery has three parts. The Laguerre cell of a generator $(x,m)$ is the set of points $z$ whose power distance $\\|z-x\\|^2 + m$ is no larger than the power distance to every other generator; the Laguerre diagram is the collection of cells with non-empty interior, and for suitable configurations it is a locally finite partition of $\\mathbb{R}^d$ into compact convex cells. Tempered configurations, defined by a bound of the form $\\sum_{(x,m)\\in\\varphi\\cap(kB^d\\times\\mathbb{R})}(1+|m|^{d+\\delta})\\le l\\,k^d$ for all $k$, provide the control of unbounded negative weights that makes the tessellation well defined. The paper also proves the measurability of the mapping that sends a weighted configuration to its Laguerre diagram, which is what allows the mixing properties of the generator to be transferred to the tessellation via $\\sigma$-algebras of cells intersecting balls.","core_discovery":"The central result, Theorem 1, states that if $\\eta$ is a stationary marked point process on $\\mathbb{R}^d \\times \\mathbb{R}$ whose typical mark $M$ satisfies $E[M_-^{(d+\\delta)/2}]<\\infty$ for some $\\delta>0$, then (a) ergodicity of $\\eta$ implies $L(\\eta)$ is almost surely a tessellation of $\\mathbb{R}^d$ and is ergodic; (b) mixing of $\\eta$ implies $L(\\eta)$ is mixing; and (c) $\\alpha$-mixing of $\\eta$ implies $L(\\eta)$ is $\\alpha$-mixing. The proof works by showing that $\\eta$ almost surely belongs to the set of tempered configurations, which controls the growth of large negative weights, and by proving that the Laguerre mapping $\\varphi\\mapsto L(\\varphi)$ is measurable, so that the tessellation's $\\sigma$-algebras can be compared with the generator's. For marked Poisson processes the moment threshold can be improved: $L(\\eta)$ is almost surely a tessellation if and only if $E[M_-^{d/2}]<\\infty$.","pith_inferences":["The strict $\\delta>0$ margin in the moment condition looks technical: for well-definedness alone the endpoint condition $E[M_-^{d/2}]<\\infty$ already suffices by a Campbell-formula argument for every stationary marked point process, and the $\\delta$ is used only in the tempered-configuration estimates behind the $\\alpha$-mixing transfer. If those estimates can be sharpened, the theorem may hold at","A similar measurable-mapping plus tempered-configuration route could transfer mixing properties to other weighted diagrams, such as power diagrams with higher-order power distances or Johnson–Mehl tessellations with random birth times.","For the Poisson case, the sharp threshold suggests a testable prediction: simulations with negative-weight tails just below the $d/2$-moment threshold should show large empty holes growing with the box, while tails just above it should tessellate.","The paper itself notes that no rate of decay of the mixing coefficients follows from the proof; closing that gap would require distributional control of the random tempered-configuration index $l(\\hat\\eta)$, not just its almost-sure finiteness."],"forward_implications":["Poisson–Laguerre tessellations are well defined exactly when $E[M_-^{d/2}]<\\infty$; when this $d/2$-moment is infinite, the origin is almost surely not covered by any cell, so the diagram is not space-filling.","For any ergodic stationary marked point process meeting the moment bound, spatial averages over the tessellation converge almost surely, so ergodic theorems apply to cell-based statistics.","Long-range independence of the generator passes to the geometry: far-apart windows of the tessellation become independent at the same qualitative rate class, mixing or $\\alpha$-mixing.","The measurability theorem makes random Laguerre tessellations legitimate random elements in a Fell-topology space, so mixing and ergodicity can be defined and checked through cells rather than through cell boundaries.","The examples show that the results cover Poisson, Cox, cluster, geostatistically marked, and some Gibbs and determinantal generators, widening the non-Poissonian theory beyond bounded weights."],"supporting_citations":[{"why":"Introduces the random Laguerre tessellation framework and supplies the regularity conditions under which a Laguerre diagram is a tessellation.","marker":"[24]"},{"why":"Provides the definition of tempered configurations used to control unbounded negative marks throughout the paper.","marker":"[29]"},{"why":"Supplies the Fell-topology and measurability background, the ergodicity criterion, and the approximation argument for mixing of tessellations.","marker":"[31]"},{"why":"Gives the definitions of ergodicity and mixing for point processes and the ergodic theorem used to show that the generator almost surely lies in a tempered configuration.","marker":"[5]"},{"why":"Provides the moment-based argument for Poisson–Laguerre tessellations with unbounded weights that the paper adapts for Theorem 1.","marker":"[26]"},{"why":"Establishes Poisson–Laguerre well-definedness near the optimal moment condition, cited for Lemma 17.","marker":"[6]"}],"fun_headline_variants":["Ergodicity, mixing, α-mixing pass from generator to Laguerre tessellation","Random Laguerre tessellations inherit all mixing properties from generator","A single moment condition transfers mixing to Laguerre tessellations","Mixing behavior of point process carries to its Laguerre tessellation","Laguerre tessellation mixing follows from generator's mixing"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The theorem requires the typical negative weight to have a moment of order $(d+\\delta)/2$ for some strictly positive $\\delta$; if the negative weights are only integrable at the critical order $d/2$, or not at all, the proof's control of far-away cells and the $\\alpha$-mixing transfer does not go through.","fun_headline_variants_meta":{"raw":{"variants":["Ergodicity, mixing, α-mixing pass from generator to Laguerre tessellation","Random Laguerre tessellations inherit all mixing properties from generator","A single moment condition transfers mixing to Laguerre tessellations","Mixing behavior of point process carries to its Laguerre tessellation","Laguerre tessellation mixing follows from generator's mixing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000519,"raw_usage":{"total_tokens":2488,"prompt_tokens":891,"completion_tokens":1597,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":1504}},"tokens_in":507,"tokens_out":1597,"duration_ms":12586,"temperature":1.0,"reasoning_tokens":1504,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:44:11.932583+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the $\\alpha$-mixing transfer at the endpoint: take an $\\alpha$-mixing marked point process whose typical mark has density proportional to $|m|^{-1-d/2}/\\log^2 |m|$ for $m\\le -e$. This mark distribution satisfies $E[M_-^{d/2}]<\\infty$ but $E[M_-^{(d+\\delta)/2}]=\\infty$ for every $\\delta>0$. If simulated Laguerre tessellations from such generators still have $\\alpha$-mixing coefficients tending to zero, the strict $\\delta>0$ in Theorem 1 is an artifact of the proof; if not, it is essential.","supporting_citations":[{"cited_title":"Roelly and A","cited_arxiv_id":null,"evidence_quote":"Provides the definition of tempered configurations used to control unbounded negative marks throughout the paper."},{"cited_title":"Schneider and W","cited_arxiv_id":null,"evidence_quote":"Supplies the Fell-topology and measurability background, the ergodicity criterion, and the approximation argument for mixing of tessellations."},{"cited_title":"Daley and D","cited_arxiv_id":null,"evidence_quote":"Gives the definitions of ergodicity and mixing for point processes and the ergodic theorem used to show that the generator almost surely lies in a tempered configuration."},{"cited_title":"Petr´ akov´ a and Z","cited_arxiv_id":null,"evidence_quote":"Provides the moment-based argument for Poisson–Laguerre tessellations with unbounded weights that the paper adapts for Theorem 1."},{"cited_title":"Gusakova and M","cited_arxiv_id":null,"evidence_quote":"Establishes Poisson–Laguerre well-definedness near the optimal moment condition, cited for Lemma 17."}],"review_version":1}