{"id":"2dbcd5e5-04fc-4a2c-b163-42f5f3051245","arxiv_id":"2501.08859","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"A curl-free constraint on a degenerate spin manifold produces a d-dimensional algebraic spin liquid, the Ampère phase, distinct from the Coulomb phase.","lead":"The paper introduces a new class of magnetic spin liquids, the Ampère phase, governed by a curl-free local constraint instead of the divergence-free rule of spin ice. It derives the power-law decay of correlations and demonstrates the phase in 2D and 3D models, extending the Maxwell analogy for frustrated magnets.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central d-exponent result rests on an untested Gaussian/i.i.d. ansatz, and the 3D correlations contain a constant offset, so the unqualified algebraic-spin-liquid claim is not established.","rationale":"The reader's weakest assumption correctly identifies the i.i.d./Gaussian ansatz as the load-bearing point of the analytical derivation. My stress-test sharpens this concern by noting that the 3D data themselves contain a fitted constant c, so the empirical claim is not actually that C(r) tends to zero; the d-algebraic component is obtained after subtracting a constant offset. The sector-restricted sampling further means the Gaussian statistics are not demonstrated for the full manifold. These points do not disprove the analytical result within the stated ansatz, but they reduce the strength of the unqualified claim. Since the reader already assigned CONDITIONAL and these considerations support that grading rather than overturning it, the appropriate recommendation is UNCHANGED. A direct high-order cumulant check on existing data would settle whether the Gaussian ansatz holds in the simulated model.","tokens_in":10186,"tokens_out":13297,"duration_ms":156087,"concrete_test":"Re-run the 3D simulation with a global Monte Carlo update that changes the all-in/all-out tetrahedron count (or use exact enumeration on the smallest pyrochlore cluster) to average over all topological sectors; in the same data, compute the non-Gaussianity parameter kappa = <F^4>/(3<F^2>^2)-1 for the coarse-grained field F on cells of several sizes. If kappa departs from zero by more than the statistical error, the i.i.d./Gaussian ansatz is invalid. If c remains nonzero after full sector averaging, the model realizes a fragmented state, not a pure d-algebraic spin liquid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof in Section III depends on the assumption that the elementary random variables are i.i.d., so that the coarse-grained field F(r) is centered Gaussian (Eqs. 2-4). But the curl-free constraint directly couples these variables; in a generic frustrated magnet the constrained measure is not guaranteed to be Gaussian. The universal exponent d in Eq. 10 is obtained by projecting the Gaussian covariance onto the longitudinal channel; without the Gaussian ansatz there is no argument that the exponent remains d. The Monte Carlo support is also weaker than the abstract suggests: in the 3D pyrochlore case (Table I), the best fit is C(r)=c+a/r^b with c=0.0028+/-0.0004, so the correlations do not decay to zero. The algebraic term is a subleading correction to a fragmented constant. In addition, the 3D cluster dynamics conserves the number of all-in/all-out tetrahedra, so the simulation probes a fixed sector and the thermal average over all sectors is not performed; the constant c may be sector-specific. The strongest claim should therefore be qualified as 'within a pure Gaussian curl-free sector' unless the ansatz is validated and the sector issue is resolved.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a new class of algebraic spin liquids, the Ampère phase, in which the local constraint is a zero-curl condition on the coarse-grained magnetization field rather than the zero-divergence condition of Coulomb phases. Section III generalizes Henley's Gaussian coarse-graining argument to derive that curl-free constraints yield algebraic magnetic correlations with exponent equal to the space dimension d (Eq. 10). The authors then construct a 2D square-lattice Ising model whose low-energy manifold is curl-free and argue a one-to-one correspondence with the square-ice Coulomb phase, confirming identical thermodynamics by Monte Carlo. In 3D, they simulate a pyrochlore-lattice vertex model with a curl-free constraint using an all-in/all-out tetrahedron cluster dynamics, reporting complementary structure factors and spin correlations fitted to c + a/r^b with b = 3.018 ± 0.030, close to d = 3. The constant c is attributed to fragmentation. The paper frames the Ampère phase as a distinct algebraic spin liquid with vectorial topological excitations (fictional current lines) instead of magnetic monopoles.","tokens_in":10384,"tokens_out":3197,"duration_ms":34655,"significance":"If the central claim is established, the Ampère phase is a conceptually new class of algebraic spin liquid that complements the well-studied Coulomb phase and extends the analogy between frustrated magnets and magnetostatics. The analytic derivation in Section III is clean under its assumptions, and the Monte Carlo fits confirming b ≈ d in 3D provide nontrivial support. The paper also offers a concrete 2D realization and identifies a specific cluster dynamics appropriate for curl-free manifolds, which is a useful methodological contribution. However, the strength of the claim as stated in the abstract and Section III is not fully supported by the 3D data, which show a constant offset, and the derivation rests on an i.i.d./Gaussian ansatz that is validated only in the specific models studied.","major_comments":[{"comment":"The 3D Ampère-phase correlation fit is reported as C(r) = c + a/r^b with c = 0.0028 ± 0.0004, so the correlations do not decay to zero. The abstract and Section III claim that correlations 'decay in space with a power law whose exponent is the space dimension d'; as stated, this is not supported by the 3D data, where the algebraic term is a correction to a finite constant. The claim should be qualified (e.g., 'up to a constant fragmented component') or an argument should be provided that this constant is a separate topological sector term that does not affect the asymptotic decay exponent.","section":"Section VI, Table I and abstract"},{"comment":"The all-in/all-out tetrahedron cluster update conserves the number of all-in and all-out tetrahedra, as the authors themselves note. The simulations prepare micro-states and probe each 'sector' separately, but no average over sectors is performed and no evidence is given that the fitted exponent or the constant c is sector-independent. Since the constant contribution could be sector-specific, the universal d-exponent claim requires either an explicit sector average or a demonstration that c and b are identical across sectors.","section":"Section VI, cluster dynamics and sector averaging"},{"comment":"The analytic derivation relies on the assumption that the elementary bricks of the vector field are independent and identically distributed, so that the coarse-grained field F(r) obeys a centered Gaussian law. However, the curl-free constraint directly couples these variables, and the constraint is imposed after the Gaussian distribution is assumed. For generic frustrated magnets the constrained measure is not guaranteed to be Gaussian, so the universal exponent d in Eq. (10) is not proven beyond this ansatz. The Monte Carlo support in two specific models is encouraging, but the manuscript should either present the result as conditional on the Gaussian/CLT assumption or provide a justification of why the CLT applies inside the constrained manifold.","section":"Section III, Eqs. (2)-(4)"}],"minor_comments":[{"comment":"Typo: 'curl-free contraint' should be 'curl-free constraint'.","section":"Abstract"},{"comment":"Typographical errors: 'corse-grained' should be 'coarse-grained', and 'vector filed' should be 'vector field'.","section":"Section III"},{"comment":"'ad-algebraic spin liquid' should be 'a d-algebraic spin liquid'.","section":"Abstract and Section II"},{"comment":"In the sentence 'we first prepared a set of micro-states', the word 'micro-states' is fine, but later 'theses sectors' should be 'these sectors'.","section":"Section VI"},{"comment":"The thermodynamic comparison in 2D (specific heat, entropy, susceptibility) is presented without error bars or an estimate of statistical uncertainty; reporting independent runs and standard errors would strengthen the claim of exact equivalence.","section":"Figure 3 and Section V"},{"comment":"The vertex-to-plaquette mapping is described verbally; a more formal statement of the spin transformation (e.g., the explicit spin relabeling or the relation between vertex types) would improve reproducibility.","section":"Section IV and Figure 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for a condensed-matter theory journal and the central idea is interesting. The main issues are load-bearing: the 3D data do not show pure algebraic decay to zero, and the derivation is conditional on an i.i.d./Gaussian ansatz that is not justified from the constrained measure. These can likely be fixed by qualifying the claims and adding sector-averaged or sector-independent evidence. The self-citation pattern is not excessive. The paper should also more clearly differentiate itself from the related work in Ref. [14]."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line first: the central derivation is correct within its stated Gaussian framework, and the 3D Monte Carlo genuinely supports the predicted exponent d, but the abstract oversells the clean power-law decay because the 3D correlations carry a constant offset. The paper applies Henley's Gaussian framework to the curl-free half of the Helmholtz decomposition: constrain F(q) to be parallel to q and the covariance becomes q_i q_j / q^2, which gives 1/r^d correlations in real space. That is the Ampère phase, the longitudinal complement of the Coulomb phase. What is actually new: the explicit Ising Hamiltonians that realize curl-free constraints (a {J1, J3} square-lattice model and a pyrochlore vertex model with all-in/all-out cluster moves), and the 3D demonstration with b = 3.018 ± 0.030 against d = 3.\n\nThe paper earns credit in a few places. The 2D mapping to square ice via the 90-degree rotation is exact, so the identical thermodynamics in Fig. 3 follow by construction and the paper says so fairly. It also cites the concurrent 2-form U(1) spin liquid preprint, acknowledges the ergodicity restriction of its cluster dynamics, and reports the constant offset in the 3D correlations. That is honest engagement, not burying.\n\nThe soft spots, in order. First, the abstract's headline claim—that correlations decay as a power law with exponent d—is not exactly what the 3D simulation shows. The fit is c + a/r^b with c = 0.0028 ± 0.0004, and the raw correlations in Table I plateau near 0.0036 at large distance. The correlations do not decay to zero; the algebraic part is a correction on top of a fragmented constant that produces the Bragg peaks the authors themselves note. The text flags this, but the abstract still oversells the clean power-law statement. Second, the all-in/all-out cluster dynamics conserves the number of such tetrahedra, so each simulation sits in a fixed sector and the fitted constant could be sector-dependent; the authors say the dynamics 'probably suffers from ergodicity breaking' but do not test whether c or b varies across sectors. Third, the derivation rests on the i.i.d./Gaussian ansatz for the coarse-grained field, and the curl-free constraint couples the very variables the ansatz treats as independent. This is the same foundational assumption used for Coulomb phases and it has historically worked well; the Monte Carlo fits with free exponents are genuine supporting evidence, but it is an ansatz, not a proof for generic models. Minor points: no error bars on the 2D thermodynamics and no code or data deposit, both easy to fix.\n\nWho this is for: theorists working on classical spin liquids and the electromagnetism analogy. It deserves a serious referee. My recommendation is conditional accept: the derivation is correct within its stated framework, the 3D model is a concrete realization, and the requested revisions are qualifications to the claims, not a rebuild. The stronger worry—that the ansatz invalidates everything—does not hold up, since the fitted exponents with free parameters are real support and the authors flag their own limitations.","headline":"A mostly correct, clean extension of Henley's framework to curl-free constraints with a genuine 3D Monte Carlo check of the d exponent, but the abstract oversells the power-law decay because the 3D correlations carry a constant offset and the sampling is sector-restricted.","tokens_in":10951,"tokens_out":6407,"would_cite":true,"duration_ms":58320,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A curl-free local constraint defines a new 'Ampère phase' whose spin correlations decay as $r^{-d}$, a distinct algebraic spin liquid in frustrated magnets.","keywords":["Ampère phase","algebraic spin liquid","curl-free constraint","Coulomb phase","frustrated magnetism","spin ice","pyrochlore lattice","cooperative paramagnet"],"falsifier":"Take a model with only a local curl-free constraint on a macroscopically degenerate manifold and compute the real-space equal-time spin correlation along a lattice direction: the paper predicts $C(r) \\sim a r^{-d}$ (plus an optional constant in 3D), so a measured exponent differing from $d$ by more than the statistical error, or a structure factor that is not dominated by the longitudinal component $q^\\mu q^\\nu/q^2$, would falsify the claim.","tokens_in":9951,"feed_emoji":"🧲","tokens_out":9107,"duration_ms":87595,"temperature":0.7,"pith_summary":"The paper proposes a new class of classical spin liquid, the Ampère phase, defined by a local curl-free condition on the coarse-grained magnetization field instead of the divergence-free (Gauss) condition that defines Coulomb phases such as spin ice. It argues that in any macroscopically degenerate ground-state manifold with cooperative-paramagnet statistics, this curl-free constraint forces spin-spin correlations to decay algebraically with distance, with exponent equal to the space dimension $d$: the phase is a $d$-algebraic spin liquid. Because the constraint is an Ampère law rather than a Gauss law, the elementary excitations are not magnetic monopoles but vectorial magnetic loops, i.e., fictional current lines associated with sources of magnetization curl. The authors demonstrate the physics with Monte Carlo simulations on a 2D square-lattice Ising model and a 3D pyrochlore vertex model, showing that thermodynamic properties match the Coulomb counterparts while the magnetic structure factors are complementary. A sympathetic reader would care because this extends the electromagnetism analogy for frustrated magnets and predicts a distinct, experimentally distinguishable form of algebraic disorder.","feed_headline":"Curl-free magnets make a new spin liquid with power-law decay","feed_subtitle":"Ampère phases extend spin ice: excitations are current lines, not monopoles, and correlations follow a power law.","key_machinery":"The load-bearing object is the coarse-grained magnetization vector field $\\mathbf{F}(\\mathbf{r})$ built from large cells of elementary spin variables, assumed independent and identically distributed and symmetric, so the central limit theorem gives a centered Gaussian distribution with variance $\\sigma^2 = V/K$. An Ampère phase is characterized by the local constraint $\\nabla \\times \\mathbf{F} = 0$, whose reciprocal-space form $\\mathbf{q} \\times \\mathbf{F}(\\mathbf{q}) = \\mathbf{0}$ forces $\\mathbf{F}$ to lie along $\\mathbf{q}$; substituting this into the Gaussian free energy $F/k_B T \\propto (K/2)\\sum_{\\mathbf{q}} |\\mathbf{F}(\\mathbf{q})|^2$ yields the longitudinal correlation tensor $q^\\mu q^\\nu/q^2$. This longitudinal projector is the exact complement of the Coulomb phase's transverse projector $\\delta^{\\mu\\nu} - q^\\mu q^\\nu/q^2$, and it is the identity that carries the argument: it produces algebraic real-space decay $r^{-d}$, complementary pinch points in the structure factor, and the current-line interpretation of defects.","core_discovery":"On its own terms, the paper's central discovery is that the curl-free constraint $\\nabla \\times \\mathbf{F}(\\mathbf{r}) = 0$ on the coarse-grained magnetization field $\\mathbf{F}$ produces an algebraic spin liquid whose reciprocal-space correlations are the longitudinal projector $\\langle F^\\mu(\\mathbf{q}) F^\\nu(\\mathbf{q})\\rangle = (1/K)\\, q^\\mu q^\\nu/q^2$, the exact complement of the Coulomb phase's transverse projector. From this identity the paper derives real-space decay $\\sim r^{-d}$ in dimension $d$, valid in any dimension, and identifies the topological defects as sources of magnetization curl: vectorial excitations interpreted as fictional current lines via Ampère's theorem. The paper constructs a 2D Ising Hamiltonian (nearest-neighbor $J_1$ plus third-neighbor $J_3$ with $J_1 = -J_3$) whose low-energy manifold is curl-free and, through a spin-pair translation, thermodynamically equivalent to square ice; Monte Carlo confirms complementary structure factors and matching specific heat, residual entropy, and $1/T$ susceptibility. In three dimensions it implements the curl-free constraint in a pyrochlore vertex model and uses an all-in/all-out cluster dynamics, finding algebraic spin correlations with exponent $b = 3.018 \\pm 0.030$, consistent with $d = 3$, together with a constant contribution that produces emerging Bragg peaks, evidence that Ampère phases, like Coulomb phases, can fragment.","pith_inferences":["The paper does not pursue this, but artificial spin-ice arrays whose lattice geometry enforces a curl-free plaquette constraint could image the fictional current lines directly in real space, making the phase visible rather than inferred from scattering.","If the claim transfers to continuous-spin or quantum models, the Ampère phase should display a distinct gapless longitudinal excitation mode rather than the transverse 'photon' modes of Coulomb spin liquids; neutron or electron spin resonance spectra would distinguish them.","The paper's decomposition of the paramagnet into divergence-free and curl-free components suggests that a complete measurement of the correlation tensor in any constrained cooperative paramagnet could be analyzed by projecting onto the two complementary tensors, potentially revealing mixed phases in lattices where the two constraints cannot be mapped by rotation."],"forward_implications":["The curl-free manifold is a $d$-algebraic spin liquid: spin correlations decay as $r^{-d}$ in $d$ dimensions, giving a concrete scattering signature.","The magnetic structure factor of an Ampère phase is the complement of a Coulomb phase; in a pure paramagnet the two components add up to a constant.","Defects are vectorial current lines rather than scalar monopoles, so magnetization-curl sources replace monopole charges and relax through contractible and non-contractible pairs at different time scales.","A 2D square-lattice model with $J_1 = -J_3$ realizes the Ampère phase with the same thermodynamics as square ice (residual entropy near $0.22$ per site), so artificial spin systems can be calibrated against known square-ice behavior.","In the 3D pyrochlore case the Ampère phase fragments, adding emergent Bragg peaks on top of algebraic correlations; existing classifications of classical spin liquids should include this state."],"supporting_citations":[{"why":"Supplies the entropy-driven free-energy and Gaussian coarse-graining formalism for Coulomb phases that this paper extends to curl-free constraints.","marker":"[12]"},{"why":"Provides the power-law spin-correlation result for divergence-free pyrochlore magnets, the Coulomb-phase baseline the Ampère phase is compared against.","marker":"[13]"},{"why":"Gives the exact residual entropy of square ice used as the benchmark for the 2D thermodynamic mapping between Coulomb and Ampère models.","marker":"[17]"},{"why":"Establishes artificial square ice as a physical realization of the Coulomb phase, the foil for the Ampère realization.","marker":"[16]"},{"why":"Introduces magnetic-moment fragmentation, which the paper uses to interpret the constant correlation offset and emergent Bragg peaks in the 3D Ampère phase.","marker":"[19]"},{"why":"Concurrent work on 2-form U(1) spin liquids that the paper positions itself against; used to state the reciprocal-space exponent result as its distinctive contribution.","marker":"[14]"}],"fun_headline_variants":["Curl-free magnets reveal a new spin liquid phase","Ampère phase: spin liquids from current lines, not monopoles","New spin liquid: curl-free constraint gives Ampère phase","Magnetic loops power new algebraic spin liquid","Spin liquid with Ampère law: no monopoles, just loops"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes the elementary spin variables in the ground-state manifold are independent and identically distributed, so the coarse-grained vector field obeys a centered Gaussian law; if microscopic correlations break this assumption, the predicted $r^{-d}$ decay need not hold.","fun_headline_variants_meta":{"raw":{"variants":["Curl-free magnets reveal a new spin liquid phase","Ampère phase: spin liquids from current lines, not monopoles","New spin liquid: curl-free constraint gives Ampère phase","Magnetic loops power new algebraic spin liquid","Spin liquid with Ampère law: no monopoles, just loops"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000216,"raw_usage":{"total_tokens":1485,"prompt_tokens":1048,"completion_tokens":437,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":664,"completion_tokens_details":{"reasoning_tokens":355}},"tokens_in":664,"tokens_out":437,"duration_ms":4464,"temperature":1.0,"reasoning_tokens":355,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:16:43.854552+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a model with only a local curl-free constraint on a macroscopically degenerate manifold and compute the real-space equal-time spin correlation along a lattice direction: the paper predicts $C(r) \\sim a r^{-d}$ (plus an optional constant in 3D), so a measured exponent differing from $d$ by more than the statistical error, or a structure factor that is not dominated by the longitudinal component $q^\\mu q^\\nu/q^2$, would falsify the claim.","supporting_citations":[{"cited_title":"Castelnovo, R","cited_arxiv_id":null,"evidence_quote":"Supplies the entropy-driven free-energy and Gaussian coarse-graining formalism for Coulomb phases that this paper extends to curl-free constraints."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the power-law spin-correlation result for divergence-free pyrochlore magnets, the Coulomb-phase baseline the Ampère phase is compared against."},{"cited_title":"Perrin, B","cited_arxiv_id":null,"evidence_quote":"Gives the exact residual entropy of square ice used as the benchmark for the 2D thermodynamic mapping between Coulomb and Ampère models."},{"cited_title":"Never- theless, we emphasize that this is related to the nature of the degree of freedom, and not to the 7 space dimensionality","cited_arxiv_id":null,"evidence_quote":"Establishes artificial square ice as a physical realization of the Coulomb phase, the foil for the Ampère realization."},{"cited_title":"Castelnovo, R","cited_arxiv_id":null,"evidence_quote":"Introduces magnetic-moment fragmentation, which the paper uses to interpret the constant correlation offset and emergent Bragg peaks in the 3D Ampère phase."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Concurrent work on 2-form U(1) spin liquids that the paper positions itself against; used to state the reciprocal-space exponent result as its distinctive contribution."}],"review_version":1}