{"id":"62350faa-69fe-44ab-87dc-b70b7361b943","arxiv_id":"2608.08396","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":3,"one_line_summary":"The authors derive explicit new analyticity bounds for the Ising free energy in three regimes, but the high-temperature proof rests on a false bipartiteness claim.","lead":"This paper claims improved sufficient conditions for analyticity of the Ising model free energy in strong-field, high-temperature, and low-temperature regimes, using the Fernandez-Procacci cluster expansion criterion with generating-function and graph-theoretic bounds. A smart generalist might read it to see how modern combinatorial enumeration is being applied to rigorous statistical mechanics, but one load-bearing graph-theoretic step appears false.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The high-temperature polymer representation (2.32) omits even subgraphs whose primitive cycles share vertices, so it is not equivalent to the high-temperature expansion; Theorem 2.5 is unsupported.","rationale":"The reader's rejection is correct, and I agree that the bipartiteness assumption in Lemma 3.2 is false under the L-infinity interaction (2.2), which creates triangular cycles for d>=2. However, the more load-bearing defect is the transition from (2.29) to (2.32): the polymer gas of primitive polygons with compatibility d>=2 cannot represent even subgraphs whose primitive components share vertices. This failure is independent of bipartiteness and occurs already on the bipartite square lattice with standard nearest-neighbour edges, so repairing Lemma 3.2 by replacing the interaction would not save Theorem 2.5. The high-temperature section is the main advertised contribution, and its central identity is invalid. The strong-field and low-temperature sections are not affected by this particular objection, but the central high-temperature claim is unsupported. The reader's REJECT verdict should stand.","tokens_in":23303,"tokens_out":7756,"duration_ms":84996,"concrete_test":"On a finite square lattice, e.g. a 3x3 or 4x4 box with free boundary conditions and standard L1 nearest-neighbour edges, compute the coefficient of x^8 in the original high-temperature expansion (2.29) with x=tanh beta, and compare it with the coefficient of x^8 obtained from the polymer sum (2.32) using primitive square cycles and compatibility d>=2. The original expansion includes the figure-eight configuration: two unit squares sharing one vertex, contributing 1 to the x^8 coefficient. In (2.32), no compatible tuple of primitive polygons has total length 8: the two length-4 squares share a vertex and are incompatible, and an 8-edge simple polygon is a different configuration. Any mismatch in this coefficient proves that (2.32) is not a valid representation and that Theorem 2.5's analyticity condition is unjustified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central high-temperature claim rests on the identity between the high-temperature expansion (2.29) and the polymer gas (2.32). This identity is false. In (2.29), the sum runs over all even subgraphs, i.e. every finite edge set in which each vertex has even degree. A connected even subgraph need not be a simple cycle: for example, two unit squares sharing a single vertex form an 8-edge even subgraph. Definition 2.2 calls such an object non-primitive, since it splits into two primitive square cycles with no common edge but a common site. In (2.32), the two square components have distance 0, so under the compatibility relation (2.13), zeta(p_i,p_j)=0 and the tuple is forbidden. The original sum, however, includes this configuration with weight (tanh beta)^8. Thus (2.32) is not equal to the high-temperature partition function. This failure is independent of the bipartiteness issue: it already occurs on the bipartite square lattice Z^2 with ordinary L1 nearest-neighbour edges. Lemma 3.2 is also false under the stated L-infinity interaction (2.2), because that graph contains triangles for every d>=2, but the decomposition failure is even more basic. Consequently, the derivation of the convergence condition (3.37) and the analyticity region in Theorem 2.5 are not supported by a valid polymer representation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the domain of analyticity of the free energy of the Ising model in three regimes: strong external field, high temperature, and low temperature. The authors use the Fernandez--Procacci cluster-expansion criterion, generating-function bounds for the number of polymers/contours, and a graph-theoretic high-temperature expansion based on Veblen's theorem. The main advertised results are an improved strong-field threshold 2h ≥ φ_st(d), a high-temperature analyticity domain β ≤ β_N(d) that is claimed to exceed the classical Simon bound and, in d=2, to give β ≤ 0.322 versus Procacci's 0.151, and a low-temperature domain β ≥ min φ_low(m,κ) that is claimed to improve on Balister--Bollobás and Procacci. The proof strategy is clearly organized, and the Fernandez--Procacci criterion is stated in the appendix in a standard form. However, the central high-temperature identity is invalid, and the strong-field counting does not match the model defined by the stated interaction.","tokens_in":23577,"tokens_out":9315,"duration_ms":103484,"significance":"If the advertised results were correct, they would constitute a substantial quantitative improvement over existing analyticity bounds for the Ising model, and the combination of generating functions with the Fernandez--Procacci criterion would be a useful contribution. The paper also makes a genuine effort to compare constants and to present the cluster-expansion framework self-contained. Unfortunately, the high-temperature theorem, which is explicitly described as one of the main contributions, rests on a false equality between the even-subgraph expansion and a polymer gas over primitive polygons. The strong-field theorem is also not established for the model as defined, because the counting argument uses the wrong graph degree. These are load-bearing errors, not presentation issues, and they invalidate the headline quantitative claims.","major_comments":[{"comment":"The passage from the even-subgraph expansion (2.29) to the primitive-polygon gas (2.32) is invalid. A finite even subgraph need not be a disjoint union of pairwise compatible primitive polygons: on Z^2 with nearest-neighbor edges, the union of two unit squares sharing exactly one vertex is an even subgraph with eight edges, but under Definition 2.2 it decomposes into two primitive square polygons that share a site. Since d(p1,p2)=0, the compatibility factor ζ(p1,p2) in (2.32) vanishes and the tuple is forbidden, whereas the original sum (2.29) contains this graph with weight (tanhβ)^8. Thus (2.32) is not equal to the high-temperature partition function. This failure is independent of the bipartiteness issue and directly invalidates Theorem 2.5 and the claimed value β_N(2)=0.322.","section":"§2.3, Eq. (2.32); §3.2, proof of Theorem 2.4"},{"comment":"Lemma 3.2 is false for the interaction defined in (2.2). With ||i−j||∞=1, the graph contains triangles for every d≥2 (for example, the vertices (0,0), (1,0), and (0,1) in d=2 are pairwise at L∞ distance 1), so not every cycle has even length. The proof's statement that 'Z^d is hypercube graph' is also not correct for this adjacency. Consequently, the even-power restriction in the convergence condition (3.37) is unjustified, and the analyticity region in Theorem 2.5 is not established even if the representation problem identified in the previous comment were resolved.","section":"Lemma 3.2 and Eq. (3.37), §3.2"},{"comment":"The strong-field counting uses the recurrence p_{L+1}(X)=X(1+p_L(X))^{2d}, which is the correct branching bound for the standard nearest-neighbor lattice with degree 2d. Under the model defined by (2.2), however, adjacency is d(i,j)=||i−j||∞≤1, so each site has 3^d−1 neighbors. The recurrence therefore does not bound the number of connected sets for the model actually under study, and the threshold φ_st(d) in Theorem 2.2 is not justified. In addition, the comparison with Friedli--Velenik is not a comparison with the same model, since Friedli--Velenik treat the standard nearest-neighbor Ising model.","section":"§3.1, Eq. (3.21); Theorem 2.2"}],"minor_comments":[{"comment":"The pressure formula in (2.34) should contain log coshβ, not β, and the thermodynamic-limit formula in (2.38) should contain d log coshβ rather than d coshβ; the displayed expressions are missing the logarithm.","section":"Eqs. (2.34) and (2.38)"},{"comment":"The notation for the lattice edges is inconsistent: (2.2) and (2.11) use ||i−j||∞=1, while (2.27) uses |x−y|=1. Since the two choices define different models, the authors should state one convention and use it consistently.","section":"Throughout"},{"comment":"The partition function Ξ^LT_Λ is called the 'large-field polymer partition function' in the sentence before (2.52); this should be 'low-temperature polymer partition function'.","section":"§2.4, Eq. (2.52)"},{"comment":"The numerical value 1.0783 is asserted as the result of a maximization over a, but the maximizing argument is not displayed; the authors should provide the calculation or a reference for this value.","section":"Eq. (2.43)"}],"recommendation":"reject","confidential_remarks":"The high-temperature result is the advertised main contribution and the identity on which it rests is false; this is not a local fix but a failure of the proposed representation. The strong-field section also needs to be recast for the adjacency actually used in (2.2). I therefore recommend rejection, even though the low-temperature part may contain salvageable ideas."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the headline high-temperature result is not supported. The identity between (2.29) and (2.32) is false, and Lemma 3.2 is false under the interaction defined in (2.2). Two squares sharing a single vertex on the square lattice form a connected even subgraph that appears in (2.29) with weight (tanh β)^8, but the two primitive polygons are incompatible in (2.32) because their distance is zero. So (2.32) is not the high-temperature expansion. Separately, the L∞ interaction in (2.2) makes the Z^d graph non-bipartite for d≥2, so the claim that all cycles are even is wrong. Both problems are load-bearing for Theorem 2.5, so the claimed improvement over Simon and Procacci is unsupported.\n\nCredit where it's due: the strong-field section is a coherent derivation (for the L∞-interaction model) and the generating-function bound leading to φ_st(d) is plausible. The low-temperature section is dense and I couldn't fully audit the contour counting, but it follows the Balister-Bollobás framework and may be right. The authors know the literature and the Fernandez-Procacci machinery is used seriously.\n\nThe other soft spot is that the paper compares against existing results for the standard nearest-neighbor L1 Ising model, while the model is defined with L∞ edges. That is not an apples-to-apples comparison, and the notation wavers between |x−y|=1 and ||i−j||∞=1 in ways that matter.\n\nWho is this for? Someone working on cluster-expansion criteria and interested in why a Veblen-based polymer representation needs a careful compatibility relation. The paper deserves a serious referee because the flaw is instructive, and the strong-field and low-temperature parts may be salvageable. But as it stands, the central claim is invalid. I would not cite it, and I'd probably not bring it to reading group except to illustrate how polymer decompositions can fail.","headline":"The high-temperature theorem fails on two independent grounds—the polymer gas (2.32) is not the high-temperature expansion, and Lemma 3.2 contradicts the stated interaction—but the strong-field and low-temperature sections have salvageable content.","tokens_in":24124,"tokens_out":9246,"would_cite":false,"duration_ms":99834,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B20","82B05"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Ising pressure is analytic for β≤0.322 in two dimensions, the paper claims, using a cluster expansion built on cycle decompositions.","keywords":["Ising model","free energy analyticity","cluster expansion","Fernandez-Procacci criterion","Veblen's theorem","generating functions","Peierls contours","self-avoiding polygons"],"falsifier":"Look for a triangle in the lattice graph defined by ||i−j||_∞=1: in d=2 the three sites (0,0), (1,0), and (0,1) are pairwise adjacent, so the graph contains an odd cycle. Since Lemma 3.2 asserts every cycle—and hence every primitive polygon—has an even number of edges, this single configuration disproves the lemma and removes the justification for summing only even powers in the high-temperature convergence condition.","tokens_in":23042,"feed_emoji":"🧲","tokens_out":5311,"duration_ms":52179,"temperature":0.7,"pith_summary":"The paper claims that for the Ising model on Z^d with an external field, the free-energy function is analytic on strictly larger parameter regions than previously known: a strong-field region, a high-temperature region at zero field, and a low-temperature region. The mechanism is a cluster expansion for polymer and contour representations, with convergence controlled by the Fernandez–Procacci criterion. The headline claim is the high-temperature region: in dimension two the pressure p^∅(β) is analytic for β≤0.322, well beyond the classical published bound and the value 0.151 previously reported for d=2. The authors obtain the improvement by rewriting the high-temperature expansion using Veblen's theorem, which decomposes even subgraphs into edge-disjoint cycles, and by counting polygons through known self-avoiding-polygon bounds.","feed_headline":"High-temperature Ising analyticity pushed to β=0.322 in 2D","feed_subtitle":"A cluster-expansion rewrite with cycle decompositions nearly doubles the known analyticity domain for the 2D Ising pressure.","key_machinery":"The load-bearing object is the polymer and cluster expansion for the Ising partition function, converted into a convergence problem by the Fernandez–Procacci criterion: the series converges if a Gruber–Kunz condition of the form sup_x ∑_{S∋x} w(S)$e^{{a|S|}}$ ≤ e^a−1 holds. Around this, the paper uses three counting devices: a generating function p(X)=X(1+p(X))^{2d} for connected subsets in the strong-field regime; Veblen's theorem in the high-temperature regime, which turns the even-subgraph expansion into a gas of primitive cycles (polygons) with activity (tanh β)^{|p|}; and a floor-stack generating function h($X^{{1/d}}$Y)≤g(X,Y)+(d−1)\\tilde g(X,Y) for primitive contours at low temperature. The sharpened analyticity regions are obtained by optimizing the auxiliary parameter a in each Gruber–Kunz condition.","core_discovery":"On the paper's own terms, the discovery is an improved analyticity theorem for the Ising pressure in all three regimes. In a strong external field, analyticity holds whenever 2h≥φ_st(d) with φ_st(d)=(4d+1)log(1+1/(4d))+log(4d), which is asymptotically much smaller than the standard textbook threshold. At zero field and high temperature, the pressure p^∅(β) is analytic for β≤β_N(d)=$tanh^{{-1}}$[$φ_1^{{high}}$(\\bar a_{high})], computed numerically as 0.322 in d=2; at low temperature, analyticity holds for β above a threshold φ_low(m,κ) whose d=2 value is 0.8226. The proof combines the Fernandez–Procacci convergence criterion with Veblen's theorem (every finite connected graph all of whose vertices have even degree is an edge-disjoint union of simple cycles) and with generating-function bounds on the number of connected subsets, contours, and polygons.","pith_inferences":["If the high-temperature argument survives a correction for odd cycles, then counts of self-avoiding polygons with even and odd length would both enter the convergence sum; because the number of odd polygons in the sup-norm graph grows similarly (2d−1)^{k−1}, the final threshold would likely be smaller than 0.322 but still above the classical bound.","The generating-function bounds appear to transfer to any lattice with bounded coordination number; one could test the strong-field and low-temperature thresholds on the triangular lattice, where the bipartite assumption already fails.","The d=2 high-temperature value 0.322 remains below the known critical inverse temperature of the square-lattice Ising model, so the method does not locate the phase transition; a natural next step is to check whether the same cluster expansion can be pushed closer to the critical point with anisotropic weights."],"forward_implications":["In dimension two, the high-temperature analyticity threshold becomes β≤0.322 instead of β≤0.151, so the technique roughly doubles the domain.","The ratio of the new strong-field threshold to the standard one decays exponentially in d, so the gain grows with dimension in the magnetic-field regime.","The low-temperature threshold in d=2, β≥0.8226, is smaller than the previously known 0.94, so analyticity extends closer to the region where order appears.","All three regimes are treated through one framework, so the same Fernandez–Procacci plus generating-function strategy can be applied to other lattice spin systems with polymer representations."],"supporting_citations":[{"why":"Supplies the Fernandez–Procacci convergence criterion that every convergence check in the paper reduces to.","marker":"[5]"},{"why":"Veblen's theorem, used to decompose even subgraphs into edge-disjoint simple cycles in the high-temperature expansion.","marker":"[15]"},{"why":"Provides the self-avoiding-polygon bound |N_k|≤2(2d−1)^{k−1} used to estimate polygon counts.","marker":"[8]"},{"why":"Supplies the generating-function method for connected subsets and floor-stack contours that sharpens the strong-field and low-temperature estimates.","marker":"[1]"},{"why":"The classical high-temperature analyticity bound against which Theorem 2.5 is compared.","marker":"[16]"},{"why":"The lecture-note bound β≤0.151 in d=2 quoted as the comparison for Theorem 2.5.","marker":"[14]"},{"why":"The textbook polymer and contour representations and the strong-field threshold that Section 2.2 improves.","marker":"[6]"},{"why":"Provides the contour decomposition and rooted-tree structure used for low-temperature contours.","marker":"[10]"}],"fun_headline_variants":["Ising analyticity pushed to β=0.322 in 2D","Veblen's theorem widens Ising free energy analyticity","Cluster expansions nearly double Ising analyticity domain","Strong-field Ising pressure analytic under sharper bound","Low-temperature Ising analyticity improved to 0.8226"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's load-bearing step is the assertion that every primitive cycle in the lattice has even length, which is needed to reduce the convergence sum to even powers but fails for the sup-norm adjacency rule already in d=2.","fun_headline_variants_meta":{"raw":{"variants":["Ising analyticity pushed to β=0.322 in 2D","Veblen's theorem widens Ising free energy analyticity","Cluster expansions nearly double Ising analyticity domain","Strong-field Ising pressure analytic under sharper bound","Low-temperature Ising analyticity improved to 0.8226"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001756,"raw_usage":{"total_tokens":6890,"prompt_tokens":860,"completion_tokens":6030,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":476,"completion_tokens_details":{"reasoning_tokens":5945}},"tokens_in":476,"tokens_out":6030,"duration_ms":40893,"temperature":1.0,"reasoning_tokens":5945,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:39:05.624911+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a triangle in the lattice graph defined by ||i−j||_∞=1: in d=2 the three sites (0,0), (1,0), and (0,1) are pairwise adjacent, so the graph contains an odd cycle. Since Lemma 3.2 asserts every cycle—and hence every primitive polygon—has an even number of edges, this single configuration disproves the lemma and removes the justification for summing only even powers in the high-temperature convergence condition.","supporting_citations":[{"cited_title":"New bounds from an old approach","cited_arxiv_id":null,"evidence_quote":"Supplies the Fernandez–Procacci convergence criterion that every convergence check in the paper reduces to."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Veblen's theorem, used to decompose even subgraphs into edge-disjoint simple cycles in the high-temperature expansion."},{"cited_title":"(eds) Polygons, Polyominoes and Polycubes","cited_arxiv_id":null,"evidence_quote":"Provides the self-avoiding-polygon bound |N_k|≤2(2d−1)^{k−1} used to estimate polygon counts."},{"cited_title":"Communications in Mathematical Physics, 273 (2), 305-315 (2007)","cited_arxiv_id":null,"evidence_quote":"Supplies the generating-function method for connected subsets and floor-stack contours that sharpens the strong-field and low-temperature estimates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The classical high-temperature analyticity bound against which Theorem 2.5 is compared."},{"cited_title":"Cluster expansion methods in rigorous statistical mechanics","cited_arxiv_id":"2308.06380","evidence_quote":"The lecture-note bound β≤0.151 in d=2 quoted as the comparison for Theorem 2.5."},{"cited_title":"Cambridge University Press (2017)","cited_arxiv_id":null,"evidence_quote":"The textbook polymer and contour representations and the strong-field threshold that Section 2.2 improves."},{"cited_title":"L., Mazel A","cited_arxiv_id":null,"evidence_quote":"Provides the contour decomposition and rooted-tree structure used for low-temperature contours."}],"review_version":1}