{"id":"7db2fa56-dd2d-4cec-9a8e-11af5e5dc7a4","arxiv_id":"2506.10068","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Power-law interactions with exponent α between d and d+1 produce critical droplets whose size diverges as h^{-1/(α-d)}, making the metastable phase effectively stable for α below a threshold α_c.","lead":"Weak long-range interactions can make a system appear to have two stable states, even when only one truly exists. This could let researchers design practical bistability in materials and experiments that were thought to lack it.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Core scaling R_c ~ h^{-1/(alpha-d)} is derived only from sharp-circular-droplet theory and is never measured directly as a function of h; an exponent mismatch would undo the apparent-bistability mechanism.","rationale":"The reader's weakest assumption is exactly the sharp-interface, single-boundary-point droplet theory; the proposed R_c(h) test directly probes whether that assumption yields the correct scaling exponent. I agree with the CONDITIONAL verdict: the mechanism is plausible, supported by the 1D rigorous results of van Enter et al. for d=1 and by the CA numerics showing a sharp crossover, but the exponent has not been validated directly in d>=2. The conclusion's algebraic lifetime statement is a real error, but it is not load-bearing for the apparent-bistability claim; the exponential nucleation time actually strengthens the practical-bistability conclusion. The main-text sentence stating rho_E ~ ell^{-1/(alpha-d)} (in the field-theory section) also appears to be a typo; it should be ell^{-(alpha-d)} to be consistent with Appendix A and with the derived ell_c ~ h^{-1/(alpha-d)}, which further motivates a careful re-derivation. None of these issues is fatal; they are addressable with the proposed measurement and minor corrections, so the verdict remains CONDITIONAL (UNCHANGED from the reader).","tokens_in":11856,"tokens_out":15811,"duration_ms":180207,"concrete_test":"In the d=2 cellular automaton at T=0.1, fix alpha=2.5 and measure the critical droplet radius R_c for h = 0.05, 0.1, 0.2, 0.4. For each h, initialize a circular droplet of radius R in the metastable background, run at least 20 realizations for t=2000 synchronous updates, and identify the threshold R_c at which the droplet grows. Plot log10 R_c versus log10(1/h); the predicted slope is 1/(alpha-d)=2. Repeat at alpha=2.2 (predicted slope 5) and alpha=2.8 (predicted slope 1.25). If measured slopes deviate by more than 20% from prediction, Eq. (2) fails for lattice droplets and the alpha_c formula loses its quantitative basis. As a robustness check, recompute the Kac normalization N_alpha with a different short-distance cutoff (e.g., omitting nearest-neighbor terms) and verify the measured exponent is unchanged.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is the exponent 1/(alpha-d) in R_c ~ h^{-1/(alpha-d)} (Eq. 2), which underlies the apparent-bistability boundary alpha_c ~ d + log(1/h)/log(beta L/f_d). Both derivations assume a perfectly sharp, circular, uniformly polarized droplet: Appendix A balances a sharp-interface energy dE ~ R^{d-(alpha-d)} against hR^d, and Appendix B evaluates the boundary effective field n_R = -[f_d(alpha)/R]^{alpha-d} at a single surface point with a Kac normalization and a hard UV cutoff |r| >= 1. If diffuse interfaces, lattice anisotropy, or cutoff-dependent droplet cores alter the energy-radius relation, the scaling exponent could differ; R_c would then have a different h-dependence and the logarithmic drift of alpha_c toward d could disappear. The paper reports no direct measurement of R_c versus h: Fig. 3(a) uses a single h=0.3, and the CA data are at one temperature. The conclusion compounds this by stating that the metastable lifetime scales as h^{-1/(alpha-d)}, whereas the paper's own nucleation argument gives tau ~ exp(c R_c^d), an exponential in h^{-d/(alpha-d)}. The exponent is therefore load-bearing and currently rests on unvalidated mean-field-style assumptions in d=2.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies metastability in Ising-type systems with power-law interactions 1/r^alpha. For the weak long-range regime d < alpha < d+1, it derives via two complementary field-theoretic arguments that the critical droplet radius scales as R_c ~ h^{-1/(alpha-d)}, an exponent that diverges as alpha -> d+. The authors then argue that in a finite system of size L, coarsening from a mixed initial state produces stable droplets of size roughly beta L, and if R_c exceeds that size, the system appears bistable for practical purposes even though a unique stable phase exists in the thermodynamic limit. This yields an apparent critical exponent alpha_c ~ d + log(1/h)/log(beta L/f_d), which drifts logarithmically slowly with L. The claims are supported by numerical simulations of a probabilistic cellular automaton in d=2, showing a crossover in the long-time magnetization and a peak in its standard deviation that sharpen with L at alpha_c ~ 2.55 for h = 0.3.","tokens_in":12101,"tokens_out":9449,"duration_ms":104302,"significance":"If the central scaling and the apparent-bistability mechanism hold, the paper introduces a conceptually interesting distinction between mathematical and physical stability in long-range systems, with potential consequences for metastability and nucleation theory. A clear strength is that the droplet-size scaling is derived from two independent approaches (a coarse-grained field-theoretic scaling argument and an explicit sharp-interface droplet calculation), and the numerical automaton is designed to preserve detailed balance, avoiding known irreversibility artifacts. The paper also makes falsifiable predictions, such as the specific h-dependence of R_c and the logarithmic drift of alpha_c with L. However, as detailed below, several of these predictions are not directly tested, and two stated results (the coarsening energy scaling and the metastable lifetime scaling) appear incorrect as written. The overall idea is promising and the central mechanism is plausible, but the current evidence is insufficient to establish the claim convincingly.","major_comments":[{"comment":"The stated coarsening energy-density scaling is internally inconsistent. The text says that for d < alpha < d+1, rho_E ~ ell^{-1/(alpha-d)}, but balancing rho_E ~ rho_h ~ h then gives ell_c ~ h^{-(alpha-d)}, not h^{-1/(alpha-d)} as claimed. The correct Bray scaling, consistent with Appendix A where the excess droplet energy is R^{d-sigma} with sigma = alpha-d, is rho_E ~ ell^{-(alpha-d)}. As written, the derivation of the central scaling law does not follow from the stated power counting; this needs to be corrected.","section":"Field theoretical insights (main text) and Appendix A"},{"comment":"The statement that 'the lifetime of the metastable phase scales as h^{-1} for alpha > d+1, and as h^{-1/(alpha-d)} for d < alpha < d+1' is incorrect. The paper's own nucleation argument gives t ~ exp(O(R^d)), so with R_c ~ h^{-1/(alpha-d)} the lifetime is exponential in h^{-d/(alpha-d)}, not a power law. The confusion appears to conflate the critical radius scaling with the lifetime; this should be corrected or the lifetime definition clarified.","section":"Conclusion"},{"comment":"The central prediction R_c ~ h^{-1/(alpha-d)} is never directly tested. Figure 3(a) shows R_c versus alpha for a single bias h = 0.3, and no h-dependence of R_c is reported anywhere. Since this exponent is the basis of the apparent-bistability mechanism and of the expression for alpha_c, a direct measurement of R_c as a function of h for fixed alpha (and a comparison with the predicted exponent) is essential to support the claim.","section":"Numerics, Fig. 3"},{"comment":"The claim that alpha_c behaves like a genuine critical point is supported only by inspection of the crossover in the average magnetization and the peak in sigma_M for L = 50, 100, 200, 400. No finite-size scaling analysis (e.g., scaling collapse of the order parameter or the Binder cumulant) is performed, and no error bars are given. Moreover, the predicted logarithmic drift of alpha_c with L is not demonstrated; the data appear to show a sharpening without a measurable shift, but this is not quantified. Given that the mechanism relies on the slow drift, this should be explicitly verified.","section":"Numerics, Fig. 3(d)-(f)"},{"comment":"The apparent boundary alpha_c ~ d + log(1/h)/log(beta L/f_d) depends on an undetermined parameter beta, the droplet fraction from coarsening, which is neither measured nor estimated in the paper. For the numerical value alpha_c ~ 2.55 (h = 0.3, L = 50 - 400) and f_d ~ 1, the formula with beta ~ O(1) gives alpha_c ~ 2.2 - 2.3, which is outside the quoted crossover. The paper should either measure beta from the numerics, or discuss the discrepancy and the sensitivity of alpha_c to beta.","section":"Field theoretical insights (Eq. for alpha_c)"}],"minor_comments":[{"comment":"The phrase 'exponential scaling of the critical droplet size R_c ~ h^{-1/(alpha-d)}' is imprecise: this is a power law in h, with an exponent that diverges as alpha -> d+. Please rephrase to avoid confusion.","section":"Abstract and Fig. 1(c)"},{"comment":"The notation '+epsilon h R^d ... - h phi_0 R^d' is confusing because both terms are proportional to h R^d; the net bulk contribution should be defined once, with its sign, so that the balance with the surface terms is clear.","section":"Appendix A, Eq. (A4)"},{"comment":"The caption states that R_c is obtained by preparing a droplet and checking when its magnetization stays constant, but the exact criterion is not given. The footnote about the periodic array of droplets is also unclear; please clarify the simulation protocol.","section":"Fig. 3(a) and footnote [44]"},{"comment":"The treatment of the ultraviolet cutoff (restricting |r| >= 1) and the claim that the neglected corrections in the d=2 integral are subleading for R >> 1 would benefit from a more explicit justification, as the scaling result depends on the cutoff being irrelevant.","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses an interesting and timely topic in long-range interacting systems, and the core idea of apparent bistability is worth publishing if the technical issues are resolved. The two clear errors (the coarsening energy scaling and the lifetime scaling) are fixable, but the missing h-dependence measurement and the lack of finite-size scaling analysis are more substantive gaps that need to be addressed before the claims can be fully supported. The paper's scope fits the journal, but the evidence is currently not strong enough for acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"David,\n\nShort version: the central scaling R_c ~ h^{-1/(alpha-d)} for d < alpha < d+1 is derived two independent ways (field-theoretic energy balance and an explicit droplet integral) and matches the earlier d=1 results and the d=2 numerics from Horikx. The genuinely new content is the general-d derivation, the closed-form f_d(alpha) prefactors, and the framing of apparent bistability with a logarithmically drifting alpha_c. I think the mechanism is right and worth taking seriously, but the evidence is thinner than the abstract claims.\n\nWhat works: the two analytic routes agree; the droplet calculation in Appendix B is concrete and yields finite f_d for alpha -> d+; and the CA simulations show a crossover at alpha_c ~ 2.55 that sharpens as L grows from 50 to 400. For a single h = 0.3 and T = 0.1 that is not a lot of data, but the trend is consistent with the mechanism. The update rule preserves detailed balance, which is the right choice when studying metastability.\n\nThe soft spots, in order of seriousness. First, the conclusion states that the metastable lifetime scales algebraically as h^{-1/(alpha-d)}. That is wrong on the paper's own terms: nucleation is exponential in R_c^d, so tau ~ exp(c h^{-d/(alpha-d)}). The body is consistent with exponential, so this is arguably a careless sentence, but it will mislead readers. Second, the h-dependence of R_c is never directly tested. All the numerics use one bias value; the exponent -1/(alpha-d) rests on the analytic arguments and prior work. Since the entire apparent-bistability picture hangs on that exponent, a direct R_c vs h measurement at fixed alpha is the natural missing figure. Third, the claim that alpha_c behaves like a genuine critical point is supported by the sigma_M peak, not by a finite-size scaling analysis (Binder cumulant or similar). With L up to 400, the sharpening is suggestive but not conclusive. Fourth, no code or data are provided, and the quantitative boundary depends on a hand-picked beta. The sharp-interface droplet assumption could modify prefactors but I doubt it changes the exponent, because the energy-balance derivation is more general.\n\nOverall: the core concept is solid and likely correct, but the paper overreaches in the conclusion and underdelivers on direct verification. I would send it to referees: it deserves a serious look, and the referees should push for the h-sweep and a corrected conclusion. It is also a good reading-group paper: the mechanism is more interesting than the current numerical support.","headline":"A likely-correct mechanism for apparent bistability with two independent analytic derivations, but the numerics never directly test the load-bearing scaling exponent and the conclusion misstates the lifetime.","tokens_in":12665,"tokens_out":3159,"would_cite":true,"duration_ms":36298,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B26","82C20","82C27"],"pacs":["05.70.Fh","05.50.+q","64.60.Q-"],"model":"deepseek-v4-flash","headline":"Critical droplet size diverges as $R_c \\sim h^{-1/(\\alpha-d)}$, so one stable phase looks like two.","keywords":["apparent bistability","long-range interactions","critical droplet","nucleation","metastability","probabilistic cellular automaton","Kac normalization","phase coexistence"],"falsifier":"Measure $R_c$ in $d=2$ for $\\alpha=2.5$ with $h$ spanning two decades, say $0.1$ to $0.01$: if $R_c$ does not grow as $h^{-2}$, Equation (2) fails. Alternatively, run the cellular automaton at $L=10^4$ with $\\alpha$ just below the apparent $\\alpha_c$ and watch for a finite-time flip to the stable phase; if such a flip occurs on accessible timescales, the apparent bistability is merely finite-size, not practically stable.","tokens_in":11583,"feed_emoji":"⚖️","tokens_out":5889,"duration_ms":58394,"temperature":0.7,"pith_summary":"This paper argues that systems with power-law interactions $1/r^\\alpha$ and a bias field $h$ can appear bistable, showing two practically stable phases, even when only one phase is truly stable. The mechanism is a critical droplet radius $R_c \\sim h^{-1/(\\alpha-d)}$ in the weak long-range regime $d<\\alpha<d+1$, which diverges exponentially as $\\alpha$ approaches $d$ from above. For $\\alpha$ below a system-size-dependent threshold $\\alpha_c$ that drifts only logarithmically with $L$, nucleating a destabilizing droplet is so unlikely that the metastable phase survives for all practical purposes. The paper supports this with a field-theoretic droplet calculation and numerics on a two-dimensional probabilistic cellular automaton.","feed_headline":"Weak long-range forces make one stable phase look bistable.","feed_subtitle":"Nucleation of the stable phase becomes exponentially rare, so the system acts bistable in practice.","key_machinery":"The carrying object is the critical droplet radius $R_c$, the radius at which the destabilizing bias $h$ balances the curvature-induced effective boundary field $n_R=-(f_d(\\alpha)/R)^{\\alpha-d}$ created by Kac-normalized power-law interactions. The argument uses two complementary routes: a coarse-grained Landau-Ginzburg energy balance in which the long-range wall energy density scales as $\\ell^{-1/(\\alpha-d)}$, and an explicit droplet integral giving the same exponent. Both routes lead to $R_c\\sim h^{-1/(\\alpha-d)}$ and to the apparent transition at $\\alpha_c$.","core_discovery":"The central claim is that in $d<\\alpha<d+1$, the effective boundary field on a droplet of radius $R$ is $n_R=-[f_d(\\alpha)/R]^{\\alpha-d}$, so balancing it against the bias $h$ gives $R_c=f_d(\\alpha)\\,h^{-1/(\\alpha-d)}$. Because the exponent diverges as $\\alpha\\to d^+$, the critical radius can exceed the system size $L$ even for moderate $h$; any droplet that coarsening can generate is then subcritical, and the system remains in the metastable state for exponentially long times. The paper calls this apparent bistability: there is a unique stable phase in the thermodynamic limit, but for $\\alpha<\\alpha_c$ with $\\alpha_c\\approx d+\\log(1/h)/\\log(\\beta L/f_d)$, the system is bistable for all practical purposes and $\\alpha_c$ behaves like a genuine critical point, sharpening with $L$.","pith_inferences":["The same $R_c\\sim h^{-1/(\\alpha-d)}$ scaling should appear in asynchronous Glauber dynamics of a long-range Ising model; testing it requires large sizes because the predicted $\\alpha_c$ drift is logarithmic.","The mechanism could protect ordered phases in driven or prethermal systems, where bias-like fields are unavoidable, making apparent bistability a resource for time-crystalline order.","A quantitative prediction: the metastable lifetime should grow roughly as $\\exp(h^{-d/(\\alpha-d)})$, a signature that could be measured directly in a single-droplet experiment."],"forward_implications":["For any finite system size $L$ there is a range $d<\\alpha<\\alpha_c$ where the metastable state has an exponentially long lifetime, so apparent bistability is unavoidable in practice.","The apparent phase boundary $\\alpha_c$ shifts only logarithmically with system size, so even very large simulations or experiments will see a sharp-looking transition.","Coarsening from a mixed initial state cannot nucleate a supercritical droplet for $\\alpha<\\alpha_c$, so the system falls back to the metastable phase.","The distinction between mathematical and physical stability is real: weak long-range interactions produce practically indistinguishable bistability without true phase coexistence."],"supporting_citations":[{"why":"Bray's domain-growth scaling supplies the long-range coarsening energy density $\\rho_E\\sim\\ell^{-1/(\\alpha-d)}$ that yields $R_c\\sim h^{-1/(\\alpha-d)}$.","marker":"[10]"},{"why":"One-dimensional Ising results with $R_c\\sim h^{-1/(\\alpha-1)}$ for $1<\\alpha<2$ are the precedent this paper generalizes to higher dimensions.","marker":"[36, 37]"},{"why":"Numerics in $d=2$ showing $R_c$ diverges as $\\alpha\\to 2^+$ motivate the weak long-range analysis.","marker":"[38]"},{"why":"These papers show that the probabilistic cellular automaton with symmetric couplings satisfies detailed balance, justifying the model choice.","marker":"[42, 43]"},{"why":"Rikvold et al. supply the short-range kinetic-Ising droplet scaling $R_c\\sim 1/h$ that the weak long-range regime is compared against.","marker":"[9]"},{"why":"Langer's theory of metastable decay frames the connection between critical droplet size and exponentially long lifetimes.","marker":"[4]"}],"fun_headline_variants":["Apparent bistability from weak long-range interactions","Weak long-range forces make one stable phase act bistable","Practical bistability from weak long-range interactions","Exponentially rare nucleation yields apparent bistability","Weak long-range interactions create apparent bistability"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes nucleation is controlled by a single sharp, circular, uniform droplet whose boundary field is computed at one surface point and whose ultraviolet divergence is cut off at $|r|\\geq 1$; if interface fluctuations, diffuse walls, or finite-size corrections change this barrier, the divergence of $R_c$ and the apparent transition would be altered.","fun_headline_variants_meta":{"raw":{"variants":["Apparent bistability from weak long-range interactions","Weak long-range forces make one stable phase act bistable","Practical bistability from weak long-range interactions","Exponentially rare nucleation yields apparent bistability","Weak long-range interactions create apparent bistability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000947,"raw_usage":{"total_tokens":4055,"prompt_tokens":967,"completion_tokens":3088,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":583,"completion_tokens_details":{"reasoning_tokens":3017}},"tokens_in":583,"tokens_out":3088,"duration_ms":20740,"temperature":1.0,"reasoning_tokens":3017,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:35:48.888551+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure $R_c$ in $d=2$ for $\\alpha=2.5$ with $h$ spanning two decades, say $0.1$ to $0.01$: if $R_c$ does not grow as $h^{-2}$, Equation (2) fails. Alternatively, run the cellular automaton at $L=10^4$ with $\\alpha$ just below the apparent $\\alpha_c$ and watch for a finite-time flip to the stable phase; if such a flip occurs on accessible timescales, the apparent bistability is merely finite-size, not practically stable.","supporting_citations":[{"cited_title":"Bray, Domain-growth scaling in systems with long- range interactions, Physical Review E47, 3191 (1993)","cited_arxiv_id":null,"evidence_quote":"Bray's domain-growth scaling supplies the long-range coarsening energy density $\\rho_E\\sim\\ell^{-1/(\\alpha-d)}$ that yields $R_c\\sim h^{-1/(\\alpha-d)}$."},{"cited_title":"Horikx,Metastability in the Two Dimensional Long- Range Ising Model, Master’s thesis (2020)","cited_arxiv_id":null,"evidence_quote":"Numerics in $d=2$ showing $R_c$ diverges as $\\alpha\\to 2^+$ motivate the weak long-range analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Rikvold et al. supply the short-range kinetic-Ising droplet scaling $R_c\\sim 1/h$ that the weak long-range regime is compared against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Langer's theory of metastable decay frames the connection between critical droplet size and exponentially long lifetimes."}],"review_version":1}