{"id":"6841ac9f-3006-4bf0-81a0-b7473c5f5394","arxiv_id":"2511.14260","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For hard-core bosons with power-law hopping on a square lattice, the critical quasi-momentum for stable superflow vanishes at α=3 and scales as (α−3)^{1+α−3} nearby.","lead":"This paper studies whether hard-core bosons with hopping that decays as a power of distance can carry stable supercurrents on a 2D square lattice. It finds that once the decay exponent drops to α=3, stable currents become impossible and the critical momentum vanishes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The asserted long-wavelength nature of the DI onset is not shown; if the first unstable mode has finite q, the inflection-point identification and the scaling law collapse.","rationale":"The reader's weakest assumption is exactly the load-bearing point: the paper identifies the DI threshold with the single-particle inflection point without displaying the numerical confirmation that the unstable modes are long-wavelength. I agree this is the correct focus. The only analytic path to the vanishing at α=3 and to the scaling law runs through this identification; without it, the agreement in Fig. 5 is not explained. The paper does have independent support: Eq. (29) reproduces known K=0 spectra, and the numerical K_c in Fig. 3 is obtained from the direct DI condition, so this is a missing verification rather than evidence of a wrong result. The scaling-exponent mismatch is secondary: the exact balance from Eq. (32) gives Δ^{1/(1−Δ)}, which is asymptotically equivalent to Δ^{1+Δ}; the paper should state this as an asymptotic form. This does not move the verdict from CONDITIONAL; it sharpens the specific check that should be required.","tokens_in":42636,"tokens_out":13483,"duration_ms":140149,"concrete_test":"For α=3.01, 3.05, 3.1, 3.2 and n=0.5, 0.7, use Eq. (29) directly: for each K, compute max_q Im[ωα(q,K)/J] over the full Brillouin zone, find the smallest K with positive maximum as K_c, and record the onset momentum q*. Then (i) check whether |q*|a→0 as α→3+; (ii) compare K_c with the inflection point defined by ∂²γα/∂K²=0; (iii) fit log K_c vs log Δ to see whether the exponent approaches 1/(1−Δ) or 1+Δ. If |q*| does not tend to zero, the inflection-point identification and the Δ^{1+Δ} scaling need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's analytic route to the central claim—K_c vanishes at α=3 and K_c ∝ Δ^{1+Δ} near α=3—uses only the condition m*^{-1}=0, i.e., the inflection point of the single-particle band γα(K). This identification is valid only if DI is triggered by infinitesimal q. In Sec. IV.B the authors state, \"Since we numerically confirm that DI ... is caused by normal modes with long wavelength,\" but no such confirmation is shown. If the first unstable mode occurs at finite q, the threshold is determined by the q-dependent radicand in Eq. (29), not by ∂²γα/∂K², and Eq. (33) has no basis. There is also an internal issue in the scaling claim: solving the inflection-point condition from the expansion Eq. (32) gives K_c ∝ Δ^{1/(1−Δ)}, not Δ^{1+Δ}; the two are only asymptotically equivalent as Δ→0. The abstract presents the truncated form as the exact scaling result. The vanishing of K_c at α=3 may survive direct numerical evaluation of Eq. (29), but the stated mechanism and the quoted scaling law rest on the unverified long-wavelength assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes the stability of current-carrying Bose-condensed states in the hard-core Bose-Hubbard model on a square lattice with power-law hopping ∝ r^{-α}. Using the mapping to the spin-1/2 XY model and a product-state mean-field ansatz, the authors derive a closed-form excitation spectrum (Eq. 29), compute Landau and dynamical instability thresholds as functions of α and filling n, and report that the critical quasi-momentum K_c vanishes at α=3. Near α=3 they claim the scaling K_c ∝ Δ^{1+Δ} with Δ=α-3. The central qualitative message is that no stable supercurrent exists for α≤3 within this mean-field treatment, which would include the Rydberg-array case α=3.","tokens_in":42883,"tokens_out":22442,"duration_ms":214908,"significance":"If correct, the vanishing of the supercurrent stability window at α=3 is a clean and experimentally relevant result, and the analytic excitation spectrum is a useful reference for future work. The K=0 limit correctly reproduces earlier spectra, which is a valuable check. The paper is concise and the mean-field derivation is transparent. However, the advertised Δ-dependent scaling law is not derived as stated, and the identification of the dynamical-instability threshold with the single-particle inflection point rests on an unshown numerical assertion. These issues are load-bearing for the paper's headline claims, but they are fixable within the manuscript's scope.","major_comments":[{"comment":"Solving m*^{-1}=0 with the expansion in Eq. (32) does not give K_ca ∝ Δ^{1+Δ} with a constant prefactor. The inflection condition ∂²γ/∂K²=0 yields, for Δ=α-3>0, K_ca = [2A(1+Δ)Δ(1-Δ)]^{1/(1-Δ)} up to the O(η^4) terms. This is not a pure power law: the prefactor depends on Δ, and the local exponent is 1/(1-Δ)=1+Δ+O(Δ²). The abstract and introduction present K_c ∝ Δ^{1+Δ} as the exact scaling form and emphasize a Δ-dependent exponent. The statement is only true asymptotically to leading logarithmic accuracy. Please derive and state the full expression, or explicitly label the result as the leading behavior with local exponent 1+Δ.","section":"Sec. IV.C, Eq. (33)"},{"comment":"The identification of K_ca with the inflection point of the single-particle band is load-bearing for Eq. (33) and for the conclusion K_c=0 at α=3. The paper states 'Since we numerically confirm that DI ... is caused by normal modes with long wavelength', but no such confirmation is shown. Equation (29) implies that infinitesimal-q modes become dynamically unstable when γ''(K)>0 for q along the current direction, but it does not exclude a finite-q instability for smaller K. If a finite-q mode goes unstable first, the threshold is set by the q-dependent radicand, not by ∂²γ/∂K². Please provide the promised q-resolved numerical evidence (for example, Im ω as a function of q for K values just below and above threshold) or give an analytic argument that the first unstable mode always has q→0.","section":"Sec. IV.B, sentence after Eq. (31)"},{"comment":"The continuum expansion for γ_α(K) contains an apparent sign error and an inconsistent treatment of the 2π factor. At η=0, the displayed result gives γ_α(0) ≈ -2π/(α-2) < 0, whereas Eq. (25) gives γ_α(0)=β_α>0. The constant term should be +2π/(α-2). Additionally, the 2π from the angular integration is dropped in the second equality and reinserted in the third. The inflection condition is insensitive to the constant sign, so the final condition is not affected, but the equation as written cannot be verified and needs to be corrected.","section":"Eq. (32)"}],"minor_comments":[{"comment":"The proportionality constants in both analytical curves are fixed by a single data point at α=3.05. Please state the number of independent numerical points and the range over which the comparison is made; with a single normalization point the test mainly checks the shape, not the absolute value.","section":"Fig. 5"},{"comment":"Typo: 'quasi-momemta' should be 'quasi-momenta'.","section":"Sec. V"},{"comment":"The labels 'ω=5', 'ω=3' in the figure appear to denote the decay exponent α, not the frequency. Please use a consistent notation.","section":"Fig. 4"},{"comment":"It would help the reader if the small-q expansion of Eq. (29) were shown explicitly, since it makes the long-wavelength criterion γ''(K)>0 apparent and would partially replace the unshown numerical confirmation.","section":"Eq. (29)"}],"recommendation":"major_revision","confidential_remarks":"The central qualitative claim—K_c vanishes at α=3—is credible and of topical interest for Rydberg-atom arrays. The main issues are the mis-stated scaling law, the missing proof/confirmation that the first unstable mode is long-wavelength, and the sign/typo in Eq. (32). These are correctable, so I recommend major revision rather than rejection. The stress-test concern about finite-q instability is partly answerable from Eq. (29), but the manuscript itself does not provide that answer."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, what to know: this paper gives a mean-field linear stability analysis of current-carrying states for hard-core bosons with power-law hopping, and its central numerical result—that the critical quasi-momentum for both Landau and dynamical instabilities goes to zero at α=3—is credible and experimentally relevant for Rydberg-atom arrays. The weakness is that the analytic explanation of the DI threshold and the scaling law near α=3 are built on an unverified assumption and a slightly misstated derivation.\n\nWhat is genuinely new: the α-dependence of the critical quasi-momenta, the vanishing at α=3, and the claimed scaling form near that point. The derivation of the excitation spectrum, Eq. (29), is careful and reproduces the known K=0 spectra from earlier work, which gives some confidence that the mean-field machinery is under control. The stability phase diagrams in Fig. 3 are useful, and the point about the critical velocity being ill-defined at α≤3 is a nice observation.\n\nNow the soft spots, in proportion. First, the identification of the DI threshold with the inflection point of the single-particle band, m*^{-1}=0, rests on the statement that DI is caused by long-wavelength modes. That numerical confirmation is not shown anywhere. If the first unstable mode sits at finite q, then Eq. (33) has no basis and the scaling analysis collapses. This matters for the mechanism, though the central vanishing at α=3 might survive because the phase diagram itself is computed from the full spectrum.\n\nSecond, the scaling law as stated is not exactly what their expansion gives. Solving γ''=0 from Eq. (32) yields K_c ∝ Δ^{1/(1−Δ)}, not Δ^{1+Δ}. The two are asymptotically equivalent as Δ→0, and for the range plotted in Fig. 5 the difference is small, but the abstract presents the truncated form as the exact result. That should be corrected to either use the exact expression or explicitly say it is the leading asymptotic form.\n\nThird, the prefactor in Fig. 5 is fixed by the single data point at α=3.05. The authors disclose this, but it should be described as a fit, not a parameter-free prediction.\n\nNone of this makes me think the paper is wrong in its main thrust. It is a serious within-subfield result, clearly written, and the deficiencies are fixable. I would send it to a good referee, asking for the long-wavelength check and a corrected statement of the scaling law.","headline":"A credible mean-field result on the collapse of stable supercurrents at α=3; the advertised scaling law is not exactly derived and the DI long-wavelength assumption is unshown, but the paper deserves refereeing.","tokens_in":43415,"tokens_out":3616,"would_cite":true,"duration_ms":36565,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes, within a mean-field treatment, that hard-core bosons with hopping decaying as r^{-α} lose all stable supercurrents at α=3, and that the critical momentum vanishes with the unusual scaling K_c ∝ (α-3)^{1+α-3}.","keywords":["hard-core bosons","long-range hopping","supercurrent stability","dynamical instability","Landau instability","XY model","mean-field theory","square lattice"],"falsifier":"Compute the excitation spectrum ω_α(q,K) numerically for α just above 3 (say, α=3.05) and locate the wavevector q at which Im ω first becomes nonzero as K increases. If that q is not arbitrarily small but finite, the inflection-point criterion is wrong. Experimentally, prepare a phase twist K in a Rydberg-atom XY simulator at α=3 and measure the lifetime of the winding; the mean-field prediction is decay for any K>0.","tokens_in":42456,"feed_emoji":"🌀","tokens_out":5396,"duration_ms":54424,"temperature":0.7,"pith_summary":"The paper studies whether a superfluid can carry a steady current when particles hop between distant sites with an amplitude decaying as a power law in distance, J(r) ∝ r^{-α}. Working with a mean-field description of the equivalent spin-1/2 XY model on a square lattice, it computes the excitation spectrum of a condensate moving with quasi-momentum K and finds the critical K beyond which the flow is Landau-unstable or dynamically unstable. The central result is that these critical quasi-momenta shrink as α decreases and reach zero at α=3, meaning that for α≤3 the Bose-condensed state cannot support a stable supercurrent. Near α=3, the dynamically unstable critical momentum obeys K_c ∝ (α-3)^{1+(α-3)}, an unusual scaling whose exponent is itself α-dependent. The α=3 case is experimentally relevant because Rydberg-atom arrays realize exactly this dipolar decay rate.","feed_headline":"Long-range hopping kills supercurrents at α=3","feed_subtitle":"Mean-field theory predicts no stable flow for hard-core bosons when hopping decays as r^{-3}, the dipolar Rydberg case.","key_machinery":"The central object is the excitation spectrum ω_α(q,K) of the Bose-condensed state, obtained by linearizing the mean-field equations of motion around the steady state φ_j = -K·r_j. Its threshold for dynamical instability is set by the inflection point of the single-particle band ε_α(K) = -(1-n)J γ_α(K), where γ_α(K)=Σ_{l≠0} (a/|r_l|)^α e^{iK·r_l} is the lattice sum encoding the long-range hopping. The paper evaluates γ_α(K) in the continuum limit for small K, using a Bessel-function integral, to extract the scaling of the effective mass m* near α=3.","core_discovery":"Within the mean-field theory, the stability of the current-carrying state is governed by the convexity of the single-particle energy band ε_α(K). For a current along x, the band becomes fully convex at α=3; at that point the effective mass m* = (∂²ε_α/∂K²)^{-1} changes sign at K=0, so the sound velocity is imaginary and the condensate is dynamically unstable for any K>0. The paper derives this from the excitation spectrum ω_α(q,K) and shows analytically, using the continuum form of the lattice sum γ_α(K), that the critical momentum vanishes as K_c ∝ Δ^{1+Δ} with Δ=α-3. It also finds that the Landau-instability critical momentum vanishes at the same α, and that the group velocity at K→0 becom","pith_inferences":["A clean falsifier of the mean-field result would be a time-resolved quench experiment in a Rydberg array: prepare a condensate with a small phase twist K at α=3 and watch for decay of the winding; the model predicts even the smallest K is dynamically unstable.","The same continuum-integral argument used here should apply to other power-law exponents and geometries; whether the vanishing point shifts away from α=3 on a non-square lattice is an immediate extension.","If quantum fluctuations beyond mean field are included, a Berezinskii-Kosterlitz-Thouless-type analysis could change the conclusion at finite temperature, especially for α≥4 where mean-field order is forbidden.","The Landau-instability threshold also vanishes at α=3, so the suppression of stable flow is not an artifact of a single instability mechanism; both criteria point the same way within this approximation."],"forward_implications":["For α≤3 the mean-field theory predicts that no stable supercurrent exists: any nonzero quasi-momentum is either Landau- or dynamically unstable, and at α=3 the flow is unstable for arbitrarily small K.","In Rydberg-atom arrays, where the dipolar exchange gives α=3, this implies that persistent current states should be absent or very fragile in the ideal hard-core boson model.","In the nearest-neighbor limit α→∞ the known critical momentum K_c a=π/2 is recovered, so the result reduces to the standard optical-lattice case.","The scaling K_c ∝ Δ^{1+Δ} with Δ=α-3 shows that the critical exponent itself depends on the distance to the critical point, a signature of long-range hopping not present in short-range models.","The single-particle band picture predicts negative effective mass in the dynamically unstable region, with imaginary sound velocity, so density perturbations grow exponentially."],"fun_headline_variants":["Supercurrents die for hard-core bosons when α=3","No stable current states at α=3 for long-range hopping","Long-range hopping kills boson superflow at α=3","Hopping decay r^{-3} ends stable boson currents","Critical momentum vanishes at α=3 for hard-core bosons"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The central conclusion rests on identifying the onset of dynamical instability with the inflection point of the single-particle band ε_α(K), an identification the paper says is numerically confirmed but does not show; if the first unstable modes have finite wavelength, the scaling and the vanishing at α=3 would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Supercurrents die for hard-core bosons when α=3","No stable current states at α=3 for long-range hopping","Long-range hopping kills boson superflow at α=3","Hopping decay r^{-3} ends stable boson currents","Critical momentum vanishes at α=3 for hard-core bosons"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000195,"raw_usage":{"total_tokens":1195,"prompt_tokens":745,"completion_tokens":450,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":489,"completion_tokens_details":{"reasoning_tokens":363}},"tokens_in":489,"tokens_out":450,"duration_ms":5075,"temperature":1.0,"reasoning_tokens":363,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T21:37:59.899826+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the excitation spectrum ω_α(q,K) numerically for α just above 3 (say, α=3.05) and locate the wavevector q at which Im ω first becomes nonzero as K increases. If that q is not arbitrarily small but finite, the inflection-point criterion is wrong. Experimentally, prepare a phase twist K in a Rydberg-atom XY simulator at α=3 and measure the lifetime of the winding; the mean-field prediction is decay for any K>0.","supporting_citations":[],"review_version":1}