{"id":"5ba0be8c-f207-4104-a2fa-069a0c6b89b1","arxiv_id":"2411.16452","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The near-critical Ising interface with magnetic field scaling δ^(15/8) converges to an explicit massive SLE3 law that is absolutely continuous with respect to SLE3.","lead":"This paper proves that the boundary between plus and minus spins in the 2D Ising model at its critical temperature, with a tiny magnetic field of size δ^(15/8), converges to a random curve called massive SLE3. This is the first rigorous construction of massive SLE3, a conjectured object from 2010, and it gives an exact formula for the curve's law.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1 rests on unverified correlation convergence in SLE-slitted rough domains; the boundary estimates in Lemma 3.9 and Claims 3.12–3.15 are partially deferred and the Lipschitz-h extension is not written out.","rationale":"The paper's central claim is substantial and the proof strategy is coherent: tightness via uniform L^p bounds on the discrete Radon-Nikodym derivative, followed by explicit identification of subsequential limits through correlation scaling limits. The use of Chelkak-Hongler-Izyurov [11] and Wu's two-arm exponent [41] is natural, and the domain assumption dim_H(∂Ω)<7/4 is designed so that the SLE_3 slit boundary (dimension 11/8) is admissible. The reader's weakest assumption correctly targets the applicability of these external results to the rough slit domains; my reading agrees that this is the most load-bearing step. I do not see a fatal internal inconsistency: the slit domain Ω_t does carry a two-arc Dobrushin structure with marked points γ(t) and b, so Theorem 2.8 is not obviously misapplied. The gaps are real but patchable: the proof is written for constant h only, Lemma 3.9 and Claim 4.3 contain deferred details, and the hypotheses of [11] are not checked. These justify the CONDITIONAL verdict but do not force a reject. I therefore recommend the reader's verdict remain unchanged, with the proposed checks serving as the natural conditions for acceptance.","tokens_in":40254,"tokens_out":21358,"duration_ms":196763,"concrete_test":"Read the statement and proof of Theorem 2.8 in Chelkak-Hongler-Izyurov [11] and list the exact regularity assumptions on the domain and its discrete approximations. Then verify that, for a fixed stopping time t, the slit domains Ω_{δ,t} converge to Ω_t in the required sense, including the close-approximation condition for the tip as a degenerate prime end, and that the boundary of Ω_t (containing an SLE_3 curve) satisfies those assumptions. As a second check, write out the omitted near-boundary estimate in Lemma 3.9 for squares not touching a or b, following [21, Theorem 5.5]; confirm the claimed δ^{β/2} bound. If the true exponent is only δ^{1/4}, the Borel–Cantelli step still works, but if the estimate requires a dimension threshold smaller than 7/4, the domain assumption in Section 2.1 must be tightened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central identification of the limiting Radon-Nikodym derivative (Theorem 1.1, Eq. (1)) requires the k-point spin correlations of the critical Ising model to scale to the explicit functions f_t^{(±,k)} in domains Ω_{δ,t} that are slit by the discrete interface and converge to Ω_t = Ω\\γ([0,t]), whose boundary contains an SLE_3 curve of Hausdorff dimension 11/8. The paper invokes Theorem 2.8 from Chelkak-Hongler-Izyurov [11] and asserts it holds with no boundary smoothness, but it does not verify the precise hypotheses of [11] for these fractal-boundary slit domains. The boundary control is then obtained via Wu's two-arm exponent in Lemma 3.9 and Claims 3.12–3.15; however, Lemma 3.9's proof omits the near-boundary contribution from squares not touching a or b, explicitly deferring to [21, Theorem 5.5], and Claim 4.3 (essential for Proposition 1.4) is likewise delegated to [5] with details left to the reader. If [11] requires more than Carathéodory convergence, or if the two-arm estimates near the SLE_3 boundary are not uniform enough, the explicit formula (1) is not justified. Additionally, the proof of Theorem 1.1 is written only for a constant external field h; the statement for general bounded Lipschitz h is asserted to follow from [7] but no argument is supplied. These are gaps in the present manuscript, not demonstrated contradictions, but they are load-bearing for the theorem as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the critical planar Ising model with Dobrushin boundary conditions on a bounded simply connected domain, perturbed by an external magnetic field. The main theorem (Theorem 1.1) states that when the field strength scales as C_σ^{-1} h δ^{15/8}, the discrete interface converges in law to a massive version of SLE3 whose Radon-Nikodym derivative with respect to SLE3 is given by an explicit series involving integrals of the functions f_t^{(k)}. The paper also proves convergence to SLE3 when the perturbation is subcritical (Proposition 1.3), degeneration to a boundary arc when the perturbation is supercritical (Proposition 1.4), and a result on the large-h limit of the massive law (Proposition 1.5). The proof strategy is based on bounding the L^p norms of the discrete Radon-Nikodym derivative to obtain tightness, and then identifying the limit through convergence of spin correlations in the slit domains, with boundary contributions controlled via two-arm exponent estimates.","tokens_in":40577,"tokens_out":6430,"duration_ms":60030,"significance":"If correct, the paper establishes the first massive SLE3 as an exact scaling limit of a near-critical Ising interface, with an explicit and parameter-free expression for the limiting law. The result directly addresses a question raised by Makarov and Smirnov and provides a concrete example of a conformally covariant (rather than conformally invariant) limit. The proof has a clear and convincing overall structure: it avoids the usual martingale-observable characterization and instead uses the Radon-Nikodym derivative together with independent results on critical Ising convergence and spin correlation scaling. The paper is honest about several deferred technical steps, which is a positive feature, but those steps are load-bearing. The explicit formula for the density and the conformal covariance property are strong and original contributions.","major_comments":[{"comment":"Theorem 1.1 is stated for a bounded Lipschitz function h, but the proof of Section 3 is carried out only for a constant external field. The sentence at the start of Section 3 says the extension to Lipschitz h is straightforward and cites [7], but no argument is supplied. Since the theorem statement covers space-dependent fields, and the extension requires nontrivial verification that the discrete sums approximate the continuum integrals with the correct weights throughout the estimates of Lemma 3.10 and Lemma 3.11, this is a load-bearing gap. The full proof for Lipschitz h must be given, or the theorem must be restricted to constant h.","section":"§3, proof of Theorem 1.1"},{"comment":"The proof of Lemma 3.9 omits the contribution to E^±_n[|V(γ_n)|] from vertices in A_n(η) lying in squares other than s_a and s_b, explicitly deferring to [21, Theorem 5.5] with the phrase 'we leave the details to the reader.' This estimate is essential for the vanishing of the interface magnetization (Lemma 3.5 and Lemma 3.7) and hence for the dominated convergence argument in Proposition 3.2 and Proposition 3.3. The missing part should be written out or, if genuinely identical to [21], the reduction should be made precise enough for the reader to check.","section":"Lemma 3.9"},{"comment":"The statement of Proposition 1.4 says 'Assume ... as in Theorem 1.1', but the proof requires ∂Ω to be smooth, as noted in Section 2.1: 'the proof of Proposition 1.4 requires ∂Ω to be more regular and we will actually take it to be smooth.' This hypothesis is absent from the proposition statement. Moreover, Claim 4.3, which is the key step showing that the FK-Ising crossing event has probability close to 1, is delegated to [5, Lemmas 5 and 7] with the extension to arbitrary bounded simply connected domains left to the reader. The proof of Claim 4.3 itself mentions a crucial restriction that ∂D has Minkowski dimension 1. These issues affect the validity of Proposition 1.4 as stated and need to be resolved, either by adding the smoothness hypothesis to the proposition or by proving the claim under the stated assumptions.","section":"Proposition 1.4 and Claim 4.3"},{"comment":"The identification of the limiting density relies on applying Theorem 2.8 ([11]) to the discrete slit domains Ω_{n,t} and their continuum limit Ω_t = Ω \\ γ([0,t]). In the proof of Proposition 3.3 it is asserted that, P-almost surely, (Ω̂_{n,t}; a_{n,t}, b_n) → (Ω_t; a_t, b) in the Carathéodory topology. This is not immediate from the convergence of γ_n to γ in the curve space (T.4), and the paper does not verify the hypotheses of Theorem 2.8 for these rough domains, in particular the 'close approximation' condition for the moving boundary point a_{n,t}. Since the whole explicit formula (1) depends on having correlation convergence in slit domains whose boundary includes an SLE3 curve, this verification is load-bearing and should be supplied.","section":"§3.2, application of Theorem 2.8 to slit domains"},{"comment":"In the proof of Proposition 3.1, the convergence of δ^{15/8} E^+_δ[∑ σ_x] and δ^{15/4} E^+_δ[(∑ σ_x)^2] is used to obtain uniform L^p bounds on the Radon-Nikodym derivative. The text refers to [2] or says this follows from Lemma 3.17, but Lemma 3.17 appears only later and its proof is omitted with 'We omit their proof as it is almost identical to that of Lemma 3.11 and Lemma 3.10.' Since tightness is a foundational part of the convergence theorem, the relevant estimates should be proved in the paper or cited with precise pointers.","section":"§3.1 and Lemma 3.17"}],"minor_comments":[{"comment":"The function f_t^{(k)} in Eq. (1) appears without the ± superscript, while Proposition 3.3 uses f_t^{(±,k)}. The notation should be made consistent and the meaning of the ± superscript (Dobrushin boundary conditions) clarified at first use.","section":"Theorem 1.1, Eq. (1) and Proposition 3.3"},{"comment":"The descriptions of the topological spaces (T.2) and (T.3) are identical ('the metrizable space of continuous functions on [0,∞) with the topology of uniform convergence on compact subsets'). Presumably (T.3) refers to the driving functions; the text should be corrected to avoid confusion.","section":"§2.3.2"},{"comment":"There are numerous typos and minor errors: 'Lipchitz' for 'Lipschitz' in §3.1, 'Corrolary' for 'Corollary', 'an ym' for 'any m', 'Edwards-Sokal cooupling' in §4.2.1, 'exitst' in the proof of Proposition 1.5, and 'lattice FK-Ising cluster boundaries to CLE 16/3' in §1.2. A careful proofreading pass is needed.","section":"Various"},{"comment":"The outline says the first step of the proof of Proposition 1.3 is to show tightness, but the details are delegated to 'the same arguments as in the proof of Proposition 3.1' with 'we leave the details to the reader' in §4.1. Since Proposition 1.3 uses a different scaling factor g_1(δ), the uniform L^p bound should be written out or the similarity to Proposition 3.1 made precise.","section":"§1.2, outline"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses an important open problem and the main theorem is likely correct, but as written the manuscript contains several gaps that are load-bearing. In particular, the proof of Theorem 1.1 is only for constant h while the statement covers Lipschitz h; Lemma 3.9 omits a key boundary contribution; Proposition 1.4 lacks the smoothness assumption that the proof and Claim 4.3 require; and the application of the correlation convergence theorem to rough slit domains is not fully justified. These are fixable in a revision, but they are not merely presentational. The author's habit of deferring technical estimates to previous papers is acceptable up to a point, but for a theorem of this importance the referee would expect the key estimates to be present or at least accompanied by precise reductions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The three things to know: this is the first massive SLE3, the machinery is original and mostly honest, and the theorem as stated is not fully proved yet. The proof works for constant h; the bounded Lipschitz extension is asserted but not written out. That is the main gap.\n\nWhat is actually new: Papon constructs a massive SLE3 as a scaling limit of the critical Ising interface with a magnetic field. Massive SLE2 and SLE4 existed; SLE3 is new. The Radon-Nikodym density expressed in terms of limits of k-point spin correlations is explicit, and the proof avoids a martingale characterization. The phase portrait (subcritical goes to SLE3, critical gives P_h, supercritical collapses to boundary arc) is a nice structural result. Tightness via L^p bounds on the discrete Radon-Nikodym derivatives is clean, and the circularity burden is low: the continuum functions come from Chelkak-Hongler-Izyurov and the critical convergence from Chelkak et al., not from this paper's own outputs.\n\nSoft spots, in proportion. First, the constant-h vs Lipschitz-h gap. Section 3 opens saying the proof is for constant h and the extension to Lipschitz h follows from [7]; Proposition 3.1 does actually prove tightness for Lipschitz h, but the identification of the limiting law (Proposition 3.2, Lemmas 3.6 and 3.8) is written with scalar h. Theorem 1.1 states Lipschitz h. That is load-bearing. A referee should ask for the extension to be written out or the theorem restricted to constant h.\n\nSecond, Claim 4.3, which is essential for Proposition 1.4, is delegated to Camia-Jiang-Newman with details left to the reader. That is a real gap in the submitted proof. Third, Lemma 3.9 defers part of the boundary estimate to [21, Theorem 5.5]. Probably fillable, but not in the manuscript. Fourth, the application of Theorem 2.8 to the random slit domains: the author claims [11] has no boundary smoothness assumption. If that is accurate, the application is fine, but the hypotheses on prime ends and the discrete approximations of the slit domains are not checked in detail. Worth a careful look in review.\n\nBottom line: the main structural argument is coherent and the result is significant. As written, I would want major revision before accepting. But a serious editor should definitely send this to referees; it is not a desk reject. I would bring it to reading group once the revised version appears.","headline":"First massive SLE3 with a coherent proof that is not yet fully written for the theorem as stated.","tokens_in":41088,"tokens_out":3548,"would_cite":true,"duration_ms":32044,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60K35","82B20","60J67"],"pacs":[],"model":"deepseek-v4-flash","headline":"A magnetic field turns the critical Ising interface into a massive SLE3","keywords":["Ising model","massive SLE","SLE3","spin interface scaling limit","external magnetic field","near-critical scaling","spin correlations","Dobrushin boundary conditions"],"falsifier":"Simulate the critical Ising interface with $H=C_{\\sigma}^{-1}h\\,\\delta^{15/8}$ on a fine grid and compare the crossing distribution of a small disk against the theorem's density truncated to $k=1$; a discrepancy that does not vanish as $\\delta\\to0$ would falsify the explicit formula.","tokens_in":2017,"feed_emoji":"🧲","tokens_out":4886,"duration_ms":117572,"temperature":0.7,"pith_summary":"This paper shows that adding a tiny external magnetic field to the critical planar Ising model changes the scaling limit of the interface that separates + and − spins. When the field strength is tuned to $C_{\\sigma}^{-1} h \\,\\delta^{15/8}$, the random interface converges in law to a massive SLE$_3$ curve, and the paper gives an explicit formula for its law relative to ordinary SLE$_3$. If the field is chosen weaker than this near-critical scale, the interface still converges to SLE$_3$; if it is stronger, the interface degenerates to a boundary arc. This provides the first massive SLE$_3$ scaling limit obtained directly from a lattice model.","feed_headline":"A magnetic field gives the Ising interface a massive limit","feed_subtitle":"At field strength δ^{15/8}, the interface becomes massive SLE3 with an explicit density; weaker fields keep SLE3, stronger fields flatten…","key_machinery":"The proof works with the discrete Radon-Nikodym derivative of the perturbed interface law with respect to the unperturbed one, expanded as a sum over the exponential of the discrete magnetization. Once the interface is fixed, the Markov property decomposes this observable into $k$-point spin expectations in the two components of the slit domain. The passage to the continuum uses the scaling limits of spin correlations in domains with non-smooth boundary, together with Wu's two-arm exponent to show that the boundary contribution vanishes and to control points close to the rough slit boundary. Tightness follows from uniform $L^p$ bounds on the Radon-Nikodym derivative, obtained from exponential-moment estimates for the magnetization field.","core_discovery":"The central result, Theorem 1.1, is that on a square lattice of mesh $\\delta$, the interface of the critical Ising model with Dobrushin boundary conditions and external field $H(x,\\delta)=C_{\\sigma}^{-1}h(x)\\,\\delta^{15/8}$ converges as $\\delta\\to0$ to a random curve whose law $P_h$ is absolutely continuous with respect to SLE$_3$. The Radon-Nikodym derivative on the $\\sigma$-field generated by $\\gamma([0,t])$ is $\\frac{dP_h}{dP_{\\mathrm{SLE}_3}}(\\gamma)\\big|_{\\sigma(\\gamma(s):0\\le s\\le t)} = \\frac{1}{Z_h(\\Omega)}\\sum_{k\\ge0}\\frac{1}{k!}\\int_{\\Omega_t^k} h(z_1)\\cdots h(z_k)\\, f_t^{(\\pm,k)}(z_1,\\dots,z_k)\\, dz_1\\cdots dz_k$, where $\\Omega_t=\\Omega\\setminus\\gamma([0,t])$ and $f_t^{(\\pm,k)}$ are the scaling limits of $k$-point spin correlations in the slit domain. Consequently, the limiting law is not conformally invariant but conformally covariant, with the density $h$ transforming with weight $15/8$. The paper also proves the two neighboring asymptotics: if the field is $\\delta^{15/8}g_1(\\delta)$ with $g_1(\\delta)\\to0$, the limit is ordinary SLE$_3$; if the field is $\\delta^{15/8}g_2(\\delta)$ with $g_2(\\delta)\\to\\infty$, the interface converges to the boundary arc carrying the $-$ boundary conditions.","pith_inferences":["The explicit density suggests a natural numerical test: at small but finite mesh, the contribution of the $k=1$ term, which involves the one-point spin correlation in the slit domain, should capture the leading effect of the magnetic field on the interface law.","The theorem opens the door to a coupling between the continuum magnetization field of the critical Ising model and an SLE$_3$ curve, in the spirit of the Gaussian-free-field coupling to SLE$_4$; such a coupling could transfer the explicit density to other observables.","The two-arm-exponent control used near the rough slit boundary is likely reusable in other near-critical interface problems where the limit curve has Hausdorff dimension below a threshold set by the arm exponent.","The large-$h$ concentration result suggests an intermediate scaling regime in which the curve's transversal fluctuations are governed by the competition between the Brownian motion and the drift term; measuring this drift could yield a macroscopic observable for the mass parameter."],"forward_implications":["If the magnetic field is weaker than the near-critical scale, the asymptotic law of the interface is unchanged from the critical model, namely SLE$_3$.","If the magnetic field is stronger than the near-critical scale, the interface collapses onto the boundary arc carrying the opposite spins, matching the behavior seen near critical percolation.","The limiting law $P_h$ is conformally covariant, so under a conformal map the mass parameter transforms with the $15/8$ dimension of the magnetization field.","The driving function of the limiting curve is absolutely continuous relative to the Brownian driving of SLE$_3$, and should be expressible as $\\sqrt{3}$ times Brownian motion plus an explicit drift once an additional continuity statement is established.","As $h\\to\\infty$, the law $P_h$ concentrates near the boundary arc: the probability that the curve exits an $\\eta$-neighborhood of that arc tends to zero."],"supporting_citations":[{"why":"Establishes that the unperturbed critical Ising interface converges to SLE$_3$, providing the reference law for the Radon-Nikodym derivative.","marker":"[9]"},{"why":"Provides the scaling limits of $k$-point spin correlations in slit domains with rough boundary, which define the functions $f_t^{(\\pm,k)}$.","marker":"[11]"},{"why":"Supplies the two-arm exponent used to show the vanishing boundary contribution and to control vertices near the slit boundary.","marker":"[41]"},{"why":"Gives the exponential-moment estimates for the magnetisation field used to prove uniform $L^p$ bounds and hence tightness.","marker":"[2]"},{"why":"Provides the Carathéodory convergence and curve-topology tools used to transfer convergence into the Loewner and curve sense.","marker":"[24]"},{"why":"Defines the massive SLE class of laws which the paper's limiting interface realizes for the Ising model.","marker":"[29]"}],"fun_headline_variants":["Critical Ising interface becomes massive SLE3 under δ^{15/8} field","Magnetic field twists Ising interface into massive SLE3","Ising interface: three fates under δ^{15/8} magnetic scaling","From SLE3 to boundary arc: magnetic field reshapes Ising interface","Massive SLE3 emerges in critical Ising under magnetic field"],"cache_read_input_tokens":43136,"weakest_assumption_plain":"The scaling limits of the spin correlations are assumed valid even in domains whose boundary is the growing rough SLE$_3$ curve; if these limits fail near that rough boundary, the explicit Radon-Nikodym derivative is not justified.","fun_headline_variants_meta":{"raw":{"variants":["Critical Ising interface becomes massive SLE3 under δ^{15/8} field","Magnetic field twists Ising interface into massive SLE3","Ising interface: three fates under δ^{15/8} magnetic scaling","From SLE3 to boundary arc: magnetic field reshapes Ising interface","Massive SLE3 emerges in critical Ising under magnetic field"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000244,"raw_usage":{"total_tokens":1640,"prompt_tokens":1158,"completion_tokens":482,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":774,"completion_tokens_details":{"reasoning_tokens":387}},"tokens_in":774,"tokens_out":482,"duration_ms":5213,"temperature":1.0,"reasoning_tokens":387,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:05:37.678864+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the critical Ising interface with $H=C_{\\sigma}^{-1}h\\,\\delta^{15/8}$ on a fine grid and compare the crossing distribution of a small disk against the theorem's density truncated to $k=1$; a discrepancy that does not vanish as $\\delta\\to0$ would falsify the explicit formula.","supporting_citations":[{"cited_title":"Con- vergence of Ising interfaces to Schramm’s SLE curves","cited_arxiv_id":null,"evidence_quote":"Establishes that the unperturbed critical Ising interface converges to SLE$_3$, providing the reference law for the Radon-Nikodym derivative."},{"cited_title":"Alternating arm exponents for the critical planar Ising model","cited_arxiv_id":null,"evidence_quote":"Supplies the two-arm exponent used to show the vanishing boundary contribution and to control vertices near the slit boundary."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the exponential-moment estimates for the magnetisation field used to prove uniform $L^p$ bounds and hence tightness."},{"cited_title":"Limits of conformal images and conformal images of limits for planar random curves","cited_arxiv_id":null,"evidence_quote":"Provides the Carathéodory convergence and curve-topology tools used to transfer convergence into the Loewner and curve sense."},{"cited_title":"Off-critical lattice models and masive SLEs , pages 362–371","cited_arxiv_id":null,"evidence_quote":"Defines the massive SLE class of laws which the paper's limiting interface realizes for the Ising model."}],"review_version":1}