{"id":"84a95db0-85e3-4096-8763-192a19d2aee1","arxiv_id":"2411.13478","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Active inverse-square particles in a trap melt from a crystal to a Wigner semicircle liquid at v0 = O(1) and to a bell-shaped profile at v0 = O(N), with covariance formulas derived in the weak-noise limit.","lead":"One-dimensional run-and-tumble particles with inverse-square repulsion show three density regimes as activity grows: crystal-like peaks, a Wigner semicircle liquid, then a bell shape. The paper derives the crystal-to-liquid crossover analytically from the Hessian of the equilibrium configuration and confirms it with simulations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (34), the closed covariance formula, is algebraically wrong: it disagrees with Eq. (35) and with direct summation, so the central covariance expression as printed is unreliable; Eq. (36) itself is consistent with Eq. (35).","rationale":"The reader's central worry is that Eq. (19) relies on an unproven fixed-point convergence assumption carried over from the active DBM. That worry does not survive scrutiny: the linearized dynamics in Eq. (17) has H_ISM=H_L^2, and H_L is diagonalized by Hermite polynomials with positive integer eigenvalues, so H_ISM is positive definite and every frozen-noise realization has the unique attracting fixed point v0 H_ISM^{-1} sigma. The covariance derivation from Eq. (19) to Eq. (26) is therefore internally sound at the linearized level. The genuine soft spot is a concrete algebraic error in Eq. (34), the displayed closed form for the covariance kernel c(u,v). The paper's own Eq. (35) is correct and Eq. (36) follows from it, but Eq. (34) as printed is wrong; a reader implementing the central covariance formula from Eq. (33)-(34) would obtain incorrect results. Because the correct expression is present in the same paper, this does not overturn the main conclusions, but it does mean the central formula is not stated correctly and must be fixed before the paper can be used as a reference. This justifies keeping a conditional verdict, with the condition being correction of Eq. (34) rather than proof of the frozen-noise fixed point.","tokens_in":23773,"tokens_out":24422,"duration_ms":259698,"concrete_test":"Compute c(u,v) three ways for u=1.5, v=0.5: (i) evaluate Eq. (34) as printed, (ii) evaluate Eq. (35), and (iii) evaluate partial sums of S=sum_{k=1}^{100} sin(ku)sin(kv)/k^4. Eq. (35) and the partial sum agree at 0.4737, while Eq. (34) gives 0.2089. Then re-derive c from the identity S=(1/2)[F(u-v)-F(u+v)], F(x)=pi^4/90 - pi^2 x^2/12 + pi |x|^3/12 - x^4/48, which yields c(u,v)=uv(u^2+v^2)/12 + pi^2 uv/6 - pi/24[(u+v)^3 - |u-v|^3]; this confirms the typo in Eq. (34) and shows Eq. (36) is unchanged.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Equation (34), the claimed closed form for c(u,v) in the central covariance expression Eqs. (33)-(34), is algebraically incorrect. For u>v, substituting any values into Eq. (34) does not reproduce Eq. (35), which is the piecewise expression the paper actually uses. For example, at u=1.5, v=0.5, direct summation of sin(ku)sin(kv)/k^4 gives 0.4737, Eq. (35) gives 0.4737, but Eq. (34) gives 0.2089. On the diagonal, Eq. (34) does not reduce to the variance formula Eq. (36), which is instead consistent with Eq. (35). The correct expression is c(u,v)=uv(u^2+v^2)/12 + pi^2 uv/6 - pi/24[(u+v)^3 - |u-v|^3], which can be written in the form of Eq. (34) only if the term inside the pi/4 bracket is 4pi uv rather than pi^2 uv (or if the pi^2uv term is moved outside the bracket). Consequently, a reader or any downstream code that uses Eq. (34) as printed will compute incorrect covariances, even though the final variance formula Eq. (36) and the ordered formula Eq. (35) are correct. The reader's stated weakest assumption, that frozen-noise realizations relax to a unique fixed point, is not actually load-bearing: in the linearized equation (17), H_ISM=H_L^2 with H_L eigenvalues 1,...,N, so H_ISM is positive definite and Eq. (19) follows immediately for every frozen realization; no anticipation from the active DBM is needed.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a one-dimensional gas of N run-and-tumble particles in a harmonic trap with repulsive inverse-square interactions (the active inverse-square model, ISM). Numerically, the steady-state density profile shows three regimes as activity increases: a peaked crystal-like profile, a smooth Wigner semi-circle liquid-like profile, and a bell-shaped profile at very high activity. The authors quantify these regimes via Lindemann's ratio and a distance measure, and they construct a phase diagram in the (a,1/γ) plane, with v0 = aN. The analytical core is a small-noise (v0 → 0), long-persistence (γ → 0) computation of the position covariance using the Hessian of the ISM, which is related to the DBM Hessian through H_ISM = H_L^2. In the large-N limit the covariance is expressed in terms of a scaling function C(w,z) (Eqs. 31-34), the single-particle variance is given in closed form (Eq. 36), and the inter-particle gap variance is derived both for γ = 0 and for finite γ (Eqs. 47 and 52). The theoretical predictions are compared with numerical simulations, with good agreement reported for a ≲ 0.1.","tokens_in":24070,"tokens_out":14149,"duration_ms":133571,"significance":"If correct, the paper provides an exact large-N description of density crossovers in an interacting active system and identifies the crossover scalings v0 = O(1) (crystal-to-liquid) and v0 = O(N) (liquid-to-bell). It extends the method developed for the active Dyson Brownian motion to the inverse-square potential, exploiting the remarkable relation between the two Hessians. The derivation is largely transparent, the numerical verification is thorough in the stated regime, and the results are falsifiable predictions about variances and gap statistics. The paper should be of interest to the active-matter and exactly-solvable nonequilibrium communities. The main analytical result is, however, marred by an algebraic error in the printed closed form for the covariance (Eq. 34), which must be corrected before publication.","major_comments":[{"comment":"The closed-form expression for c(u,v) is algebraically incorrect. As printed, c(u,v) = (1/6)[uv(u^2+v^2)/2 + (π/4)(|u−v|^3 − (u+v)^3 + π^2 uv)] does not agree with the piecewise form in Eq. (35) nor with the variance limit in Eq. (36). For example, at u=1.5, v=0.5, direct summation of sin(ku)sin(kv)/k^4 gives 0.4737, Eq. (35) gives 0.4737, but Eq. (34) gives 0.2089. On the diagonal, Eq. (34) does not reduce to Eq. (36), which is instead consistent with Eq. (35). The coefficient π^2uv inside the π/4 bracket should be 4πuv (equivalently, the π^2uv/6 term should appear outside the bracket). As written, a reader or downstream code using Eq. (34) will compute incorrect covariances, so this central formula must be corrected.","section":"Sec. 4.2, Eq. (34)"},{"comment":"The linearization of the equation of motion is justified by the statement that the typical fluctuations δx_i are small compared with the equilibrium gap, which scales as 1/√N. In the liquid regime where the formula is applied, v0 = O(1), the typical displacement is v0/√N, i.e., of order the gap, so this stated premise fails. The a posteriori validity check in Eq. (53) instead uses the gap fluctuation g_{i,1} ~ v0/N^{3/2}, which is indeed the correct small parameter because the interaction force depends only on differences δx_i − δx_j. The paper should restate the linearization condition consistently in terms of gap fluctuations and explain why the individual-position variance formula (36) is nevertheless valid; as written, the derivation's explicit assumption is false in the regime of interest, even though the numerical results support the final formula.","section":"Sec. 4, before Eq. (17), and Sec. 4.4"},{"comment":"The upper limit of the sum defining C(w,z) is written as N, but the subsequent closed-form evaluation and the finite-γ expression in Eq. (42) use an infinite sum. Since the approximation leading to Eq. (32) is valid only for k ≪ N, and the 1/k^4 tail is negligible, the upper limit should be ∞ (or the text should state that the sum is extended to infinity with negligible error). This is a typographical inconsistency in a displayed equation that could mislead a reader implementing the formula.","section":"Sec. 4.2, Eq. (32)"}],"minor_comments":[{"comment":"The finite-γ results are imported from the companion paper [51] without derivation. A short derivation or a more explicit statement of the assumptions used in that derivation would improve self-containedness, since these formulas are used for quantitative comparison with simulations.","section":"Sec. 4.2, Eqs. (41)-(43)"},{"comment":"The caption notes that the first transition occurs at a values that depend on N, which is correct because a = v0/N and the crossover is at v0 = O(1). A one-sentence reminder that a = v0/N would help avoid apparent tension with the statement that the crossover is independent of N in terms of v0.","section":"Sec. 3.4, Fig. 8 caption"},{"comment":"The expansion D(w,z) = (π^2/24)(w−z)^2 + O(w^3,z^3) would benefit from a brief derivation or at least a statement that the linear terms cancel by symmetry, since the reader cannot easily verify the coefficient from the preceding expressions as written.","section":"Sec. 4.3, Eq. (46)"}],"recommendation":"major_revision","confidential_remarks":"The algebraic error in Eq. (34) appears to be a typographical slip rather than a conceptual flaw, since Eq. (35) and Eq. (36) are correct and the simulations validate the final variance formula. However, the printed central covariance formula is wrong and would mislead any reader who uses it directly. The linearization-justification issue in Sec. 4 is a rigor concern that should be fixed by clarifying the small parameter. The paper is otherwise a solid contribution and, after these corrections, would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is a solid paper with one concrete fix needed. The printed closed-form covariance (Eq. 34) is algebraically wrong. It does not match the piecewise expression in Eq. (35) or direct summation; the final variance formula (Eq. 36) and the ordered covariance (Eq. 35) are correct, but Eq. (34) as written will send downstream users astray. Easy typo-level fix, but it sits in the center of the paper.\n\nWhat's new: the active inverse-square model with run-and-tumble noise is studied for the first time, and the paper derives explicit 1/N covariance and gap-variance formulas that are not in the earlier active DBM papers. The Hessian/eigenfunction method from Refs. [30,49] is applied cleanly, and the numerics corroborate the analytic formulas for v0/N up to about 0.1 across the tumbling rates. I also like the edge analysis: no wings, uniform 1/N variance scaling at the edge, and the sharp contrast with the active DBM. The crossover scalings v0 = O(1) and v0 = O(N) are simple and convincingly supported.\n\nSoft spots: first, the Eq. (34) error. I checked the algebra: the pi^2 uv term sits inside the pi/4 bracket, which cannot be right; the bracket needs (|u-v|^3 - (u+v)^3 + 4 pi uv) or alternatively the pi^2 uv term must move outside. For u=1.5, v=0.5, direct summation gives 0.4737 while Eq. (34) gives 0.2089. So the printed central formula is unreliable, even though Eq. (35) and Eq. (36) are correct. Second, the finite-gamma expressions are imported from the companion paper [51] without derivation; acceptable if that paper is solid, but the current paper should at least outline the derivation or clearly mark them as external. Third, the phase boundaries (Lindemann ratio 0.5, chi=0.1) are hand-picked; they work for the numerics shown but they're conventions rather than sharp transitions. That's fine for a crossover study, but the paper should say so plainly. The reader's worry about fixed-point convergence for frozen noise is not actually a weak point: H_ISM = H_L^2 is positive definite, so Eq. (19) follows immediately from the linearized dynamics; no assumption imported from the DBM case is needed.\n\nBottom line: this deserves a serious referee. The core derivations are standard, the numerics are honest, and the one real defect is a fixable algebraic typo. I'd recommend sending it to review with a request to correct Eq. (34) and to state more carefully what is imported from [51].","headline":"Solid active-matter paper with clean Hessian calculations, but Eq. (34) is algebraically wrong and needs a fix before this is publishable as is.","tokens_in":24712,"tokens_out":2385,"would_cite":true,"duration_ms":20510,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82C22","60B20","82C31"],"pacs":["05.40.-a","05.20.-y","64.70.-n"],"model":"deepseek-v4-flash","headline":"Run-and-tumble particles with inverse-square repulsion melt from a crystal into a liquid when their activity reaches order one, and into a bell-shaped cloud when activity reaches order N, with exact large-N fluctuation formulas setting…","keywords":["run-and-tumble particles","inverse-square interaction","Calogero-Moser model","Wigner semi-circle","Lindemann ratio","active matter","position covariance","density crossover"],"falsifier":"Integrate Eq. (2) numerically at small tumbling rate (e.g. γ = 0.01) and small activity (e.g. v0 = 0.1) for N = 128, extract the steady-state covariance matrix, and compare it element-by-element with the scaled formula v0²/N C(x_i/√(2N), x_j/√(2N)); a disagreement larger than statistical error for bulk pairs would refute the central claim.","tokens_in":23527,"feed_emoji":"🌡️","tokens_out":6489,"duration_ms":63593,"temperature":0.7,"pith_summary":"This paper studies N run-and-tumble particles in one dimension, confined by a harmonic trap and repelling one another through an inverse-square power-law potential. It attempts to establish that the steady-state density crosses over through three regimes as activity grows: sharp crystalline peaks, a smooth Wigner semicircle, and a bell-shaped cloud. The analytical centrepiece is a closed-form expression for the steady-state covariance of particle positions in the weak-noise limit, from which the single-particle variance follows. Comparing fluctuations with the interparticle spacing and with the support size yields crossover thresholds at v0 ~ O(1) and v0 ~ O(N), which the paper verifies by simulation. If correct, this provides a rare exact large-N description of a genuinely nonequilibrium interacting active system.","feed_headline":"Active-particle crystals melt at a size-independent speed","feed_subtitle":"Exact formulas locate the crystal-to-liquid crossover at v0 of order one, independent of system size.","key_machinery":"The load-bearing object is the Hessian matrix H_ISM of the inverse-square model, evaluated at the Hermite-polynomial ground state. The paper uses the identity H_ISM = H_L², where H_L is the Hessian of the logarithmic (Dyson) model, and diagonalizes H_L with Hermite-polynomial eigenvectors whose eigenvalues are the integers 1,…,N. This turns the covariance into a Chebyshev sum with a 1/k⁴ weight, which converges quickly enough for a large-N evaluation that is also valid for edge particles; the same structure yields Lindemann's ratio and the density-vs-Wigner distance measure χ used to locate the crossovers.","core_discovery":"The paper's central claim is that in the weak-noise, long-persistence limit the displacement covariance is ⟨δxiδxj⟩ = v0²/N C(w,z) with C(w,z) = c(arccos w, arccos z)/(√(1−w²)√(1−z²)), and the single-particle variance is ⟨δxi²⟩ = v0²/N V(w), V(w) = arccos² w (π − arccos w)²/[6(1−w²)], where w and z are equilibrium positions scaled by √(2N). These closed forms imply that fluctuations are of order v0/√N for bulk and edge particles alike, that the crystal-like multi-peaked density melts to a Wigner semicircle when v0 ~ O(1), and that the semicircle gives way to a bell-shaped profile when v0 ~ O(N). The same covariance yields the variance of interparticle gaps near the trap centre, and a finite-tumbling-rate extension, all corroborated by numerical simulations.","pith_inferences":["Because the same covariance machinery works for the Dyson and inverse-square models with weights 1/k² and 1/k⁴ respectively, the method likely extends to harmonically confined Riesz gases with other power-law exponents; the edge behaviour and crossover thresholds should interpolate between these two cases.","The 1/x³ tail appears independent of the interaction potential provided particles repel infinitely on contact, which suggests a testable experimental signature: trapped active colloids should show a power-law edge decay whose exponent does not depend on the interaction law.","The Lindemann-ratio criterion fixed at 0.5 is a phenomenological choice; the analytic variance-to-gap ratio provides a parameter-free observable that could be used in experiments to map the melting line in the (v0, γ) plane.","If the weak-noise fixed-point assumption fails for some realizations at finite γ, the crossover thresholds would shift; a direct check would be to compare the simulated distribution of long-time displacements with the predicted form v0 H^{-1} σ for many frozen noise realizations."],"forward_implications":["The density profile stays close to the Wigner semicircle across a broad intermediate range of activity, with deviations confined to the edges, and no extended 'wings' appear because the nearest-neighbour force at the edge is strong enough to balance the trap.","Position fluctuations decay as 1/N for both bulk and edge particles, so the small-displacement theory remains valid up to v0 ~ O(N), covering the entire crystal and liquid regimes.","The crystal-to-liquid crossover occurs at v0 ~ O(1), independent of N, while the liquid-to-bell crossover occurs at v0 ~ O(N).","Near the trap centre the variance of the gap between particles i and i+n grows as n² with coefficient π⁴/96 v0²/N³ in the long-persistence limit, and the finite-γ version matches simulations.","In the strongly active regime the density has a bell-shaped bulk and a 1/x³ tail that contains only O(1) particles, with strong sample-to-sample fluctuations."],"supporting_citations":[{"why":"Supplies the weak-noise fixed-point method and the large-N approximation used for the covariance calculation.","marker":"[30]"},{"why":"Earlier study of the active Dyson Brownian motion whose density-regime phenomenology and edge behaviour this paper compares against.","marker":"[34]"},{"why":"Gives the ground state as Hermite zeros and the identity H_ISM = H_L² used to diagonalize the Hessian.","marker":"[49]"},{"why":"Establishes the Wigner semicircle density for harmonically confined long-range repulsive particles, used as the baseline liquid profile.","marker":"[50]"},{"why":"Companion derivation of the finite-γ covariance formula that the paper adopts for its extension.","marker":"[51]"},{"why":"Provides the exact eigenvectors and integer eigenvalues of H_L used in the spectral decomposition.","marker":"[53]"}],"fun_headline_variants":["Active crystals melt at activity of order one","Crystal-to-liquid crossover at v0 ~ 1 for active particles","Exact formulas show active crystal melt at v0~1","Size-independent melt point for active crystals","Active crystal melts at v0~1 regardless of N"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes that in the γ → 0 limit every frozen realization of the run-and-tumble noise drives the linearized system to one unique fixed point, so the displacement is simply v0 times the inverse Hessian acting on the noise; if some realizations never relax, the variance formulas and the crossover thresholds do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Active crystals melt at activity of order one","Crystal-to-liquid crossover at v0 ~ 1 for active particles","Exact formulas show active crystal melt at v0~1","Size-independent melt point for active crystals","Active crystal melts at v0~1 regardless of N"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00098,"raw_usage":{"total_tokens":4146,"prompt_tokens":917,"completion_tokens":3229,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":533,"completion_tokens_details":{"reasoning_tokens":3151}},"tokens_in":533,"tokens_out":3229,"duration_ms":22783,"temperature":1.0,"reasoning_tokens":3151,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:23:46.838735+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Integrate Eq. (2) numerically at small tumbling rate (e.g. γ = 0.01) and small activity (e.g. v0 = 0.1) for N = 128, extract the steady-state covariance matrix, and compare it element-by-element with the scaled formula v0²/N C(x_i/√(2N), x_j/√(2N)); a disagreement larger than statistical error for bulk pairs would refute the central claim.","supporting_citations":[{"cited_title":"Touzo, P","cited_arxiv_id":null,"evidence_quote":"Supplies the weak-noise fixed-point method and the large-N approximation used for the covariance calculation."},{"cited_title":"Touzo, P","cited_arxiv_id":null,"evidence_quote":"Earlier study of the active Dyson Brownian motion whose density-regime phenomenology and edge behaviour this paper compares against."},{"cited_title":"Agarwal, M","cited_arxiv_id":null,"evidence_quote":"Gives the ground state as Hermite zeros and the identity H_ISM = H_L² used to diagonalize the Hessian."},{"cited_title":"Agarwal, A","cited_arxiv_id":null,"evidence_quote":"Establishes the Wigner semicircle density for harmonically confined long-range repulsive particles, used as the baseline liquid profile."},{"cited_title":"Spatio-temporal fluctuations in the passive and active Riesz gas on the circle","cited_arxiv_id":"2411.01355","evidence_quote":"Companion derivation of the finite-γ covariance formula that the paper adopts for its extension."},{"cited_title":"Ahmed, M","cited_arxiv_id":null,"evidence_quote":"Provides the exact eigenvectors and integer eigenvalues of H_L used in the spectral decomposition."}],"review_version":1}