{"id":"d52b6547-0621-4d31-9869-4479e8459f34","arxiv_id":"1908.02068","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A floating, gapless incommensurate phase occurs between the 1/3 and 1/4 crystalline lobes and around the 1/5 lobe in the 1D Rydberg Ising chain, with quantitative phase boundaries determined by tensor network calculations.","lead":"This paper maps the ground-state phases of a one-dimensional chain of interacting Rydberg atoms and finds a floating, incommensurate phase between the known crystalline and disordered phases. The result gives experimenters concrete parameter regions where this subtle phase should appear in existing Rydberg simulators.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The c=1±5% criterion used to map the floating phase is the load-bearing soft spot: because the disorder-to-floating transition is Kosterlitz-Thouless, finite-bond-dimension iMPS can mimic c=1 in a gapped near-critical region, so the reported phase boundaries are not yet quantitatively secure.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing concern: the automatic c = 1 ± 5% classification criterion may overestimate the floating phase extent near the Kosterlitz-Thouless transition, and the manuscript lacks error bars or convergence data for the phase boundaries. My stress-test pass did not reveal a different, more serious flaw. The central physical claim—that floating phases exist in the Rydberg Ising chain and occupy experimentally accessible parameter regions—is supported by the interior c ≈ 1 points, the incommensurate drifting structure-factor peaks, and consistency with earlier theoretical predictions. The issue is quantitative: the boundary locations, the claim that the 1/5 lobe is 'fully coated', and the phrase 'rather prominently' all depend on the threshold criterion. The authors' own caveat makes this an acknowledged soft spot rather than an internal inconsistency. I therefore agree with the CONDITIONAL verdict and recommend no change. The concrete test I propose—checking whether S(k_max) grows with the Fourier window R at the floating/disordered boundary—directly distinguishes a power-law-correlated floating phase from a gapped near-critical phase without relying on the disputed c-threshold, and would settle whether the overestimation is material.","tokens_in":8044,"tokens_out":3898,"duration_ms":45522,"concrete_test":"Pick three representative points on the x = 4 cut: one clearly inside the gray region, one at the gray/white boundary, and one just outside the boundary. For each, compute the structure factor S(k) maximum as a function of the Fourier window R = 10^3, 3×10^3, 10^4, 3×10^4. In a true floating phase with power-law correlations, the peak height grows as ~R^{1-α} with α ≤ 1/4; in a gapped phase it saturates. Independently, rerun the boundary-point iDMRG at χ = 128, 256, 512 and track the extracted c from S versus log ξ; if the inferred c drifts away from 1 with increasing χ, and the boundary location shifts monotonically as a function of 1/χ, the c=1±5% region overestimates the floating phase extent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is the extent and location of the floating phase in Fig. 1, and every gray boundary is set by the automatic classification criterion c = 1 ± 5% extracted from S = (c/6) log ξ + const. (Eq. 2). The authors themselves state that 'the extent of the floating phase might be slightly overestimated using this criterion' (Methods, automatic classification). This is not a minor footnote: the disorder-to-floating transition is Kosterlitz-Thouless, where the correlation length diverges exponentially in the distance to the transition. In infinite-DMRG with finite bond dimension χ, the accessible correlation length ξ is bounded, so a gapped phase sufficiently close to the KT point will have a very large but finite ξ and can produce an effective c close to 1 over a sizable parameter window. The 5% threshold then converts this near-critical region into a bona fide floating phase. The structure-factor data in Figs. 3 and 4 are consistent with incommensurate power-law peaks, but they are evaluated at finite R (up to 10000) and are interpreted qualitatively rather than used to extract the correlation exponent. The existence of at least some floating phase is well supported by the interior c ≈ 1 points, so this concern does not overturn the main prediction; it makes the quoted boundaries, the 'fully coated 1/5-lobe', and the statement that floating phases are 'rather prominent' quantitatively uncertain.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a quantitative ground-state phase diagram for the one-dimensional van der Waals Rydberg Ising chain described by Eq. (1), using infinite matrix product state (iMPS) simulations with iDMRG and the TeNPy library. The authors identify the usual gapped crystalline 1/q lobes and the disordered phase, and additionally locate extended gapless, incommensurate \"floating\" phases that interpolate between crystalline lobes and the disordered regime. The floating phase is diagnosed by extracting a central charge c from the entanglement-entropy scaling relation S = (c/6) log ξ + const. (Eq. (2)) and by tracking the parameter-dependent drift of structure-factor peaks (Eq. (3)). They conclude that floating phases are prominent in experimentally accessible parameter ranges and that the 1/5-lobe is fully coated by a floating phase. The paper is a numerical study without analytic derivations, and it includes explicit caveats about the automatic classification criterion.","tokens_in":8321,"tokens_out":3239,"duration_ms":34198,"significance":"If the reported phase diagram is quantitatively correct, this is an important result for the Rydberg-array quantum simulator community: it predicts a gapless, incommensurate phase in a parameter regime that existing experiments can reach, and it clarifies the nature of the melting transitions of the crystalline lobes. The work uses standard and independently motivated numerical diagnostics (central charge scaling and structure-factor peak drift) rather than fitting to a preconceived phase diagram, and it explicitly acknowledges the main methodological limitation. The existence of at least some floating phase is well supported by interior points with c ≈ 1 and by the smoothly drifting structure-factor peaks. The main weakness is that the quantitative extent of the floating phase, which is the paper's central quantitative claim, rests on a c = 1 ± 5% threshold that the authors themselves state may overestimate the floating-phase extent, and the phase boundaries in Fig. 1 are drawn as guides to the eye without error estimates.","major_comments":[{"comment":"The automatic classification of floating phase via the c = 1 ± 5% criterion in Eq. (2) is load-bearing for the quantitative extent of the gray regions in Fig. 1. The authors note that the extent \"might be slightly overestimated\" because the disorder-to-floating transition is Kosterlitz-Thouless, with logarithmic corrections. This caveat is stronger than \"slight\" in practice: for a KT transition the correlation length diverges exponentially, and with finite bond dimension χ the iMPS correlation length is bounded, so a gapped phase sufficiently close to the KT point can mimic c ≈ 1 over a sizable parameter window. I request a quantitative assessment: for representative parameter points near the gray boundaries, show convergence of c as χ increases, or replace/augment the c criterion with a more discriminating diagnostic (e.g., the Luttinger parameter K or the power-law exponent of the structure-factor peak) and provide error bars on the phase boundaries.","section":"Methods (automatic classification) and Fig. 1"},{"comment":"The phase transition lines in Fig. 1 are explicitly described as \"guides to the eye,\" and the transition points are said to come from \"a series of simulations,\" but the procedure for locating each transition point and the associated uncertainty are not given. Since the paper's stated goal is a \"quantitative ground state phase diagram,\" the lack of error bars or a defined crossing criterion for the dots undermines the quantitative claim. Please specify how each transition point is identified (e.g., maximum of some susceptibility, crossing of energy estimates, or drift of density imbalance) and report uncertainties, at least for representative cuts such as x = 4 and y = 3.6.","section":"Fig. 1 and phase-diagram determination"},{"comment":"The statement that the 1/5-lobe is \"fully coated\" and that \"all 1/q-lobes for q≥5 are fully immersed into the floating phase\" goes beyond the simulation data shown: Fig. 1 shows only a finite region, and the text justifies the q≥5 extrapolation by arguments in Refs. [15, 16]. This is a plausible conjecture but should be explicitly labeled as an extrapolation rather than as a numerical result. In its current form, the abstract and conclusion present this as an established part of the phase diagram, which overstates the evidence.","section":"Discussion of the 1/5-lobe and q ≥ 5 lobes"}],"minor_comments":[{"comment":"There is a typographical error: \"R = 10000 for the remaining. phases\" should read \"R = 10000 for the remaining phases.\"","section":"Footnote [39]"},{"comment":"The symbol \"Luc\" is used for the iMPS unit-cell size, but it is not typeset as a subscript; please use \"L_uc\" for readability and define it before first use.","section":"Method (unit-cell notation)"},{"comment":"The structure factor is defined up to proportionality, which is fine, but the normalization convention (including R and possible site factors) should be stated explicitly, since the comparison of peak sharpness between different R values in Figs. 3 and 4 is qualitative.","section":"Eq. (3)"},{"comment":"The structure-factor panels are interpreted qualitatively as showing power-law diverging peaks, but no power-law exponent is extracted from the finite-R data (R up to 10000 in Fig. 3 and Fig. 4). A quantitative extraction of the peak exponent as a function of x or y would strengthen the identification, even if only for one representative cut.","section":"Structure-factor interpretation"},{"comment":"The two examples in Fig. 2 are useful sanity checks (c ≈ 1 in the floating phase and c ≈ 0.5 at the Ising transition), but the figure would benefit from indicating the range of χ values used in each fit and the fit residuals, so the reader can judge the quality of the linear scaling.","section":"Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely topic and uses appropriate tensor-network methods; the main prediction of floating phases is credible. My concern is that the central quantitative claim — the extent of the floating phase in Fig. 1 — rests on a criterion that the authors admit overestimates the phase extent, and no error estimates are given for the phase boundaries. This is fixable within the scope of the manuscript by adding convergence checks, alternative diagnostics, and uncertainty quantification, so I recommend major revision rather than rejection. I would also encourage the authors to make the extrapolation to q ≥ 5 explicit and clearly separated from the direct numerical results."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is the first quantitative phase diagram of the van der Waals Rydberg chain that includes the floating phase, and the central identification is credible. The paper deserves a real referee. But the plotted phase boundaries are less secure than the figure suggests, and I would want that addressed before quoting numbers.\n\nWhat's new: Weimer-Buchler and Sela et al. predicted floating phases in this class of models, and hard-boson/dipolar studies saw slivers, but nobody had mapped them in the actual 1/r^6 Rydberg chain. The authors do it with iMPS, using central charge from entanglement scaling plus structure factor peak drift. The interior of the floating phase is supported by c~1 points and drifting incommensurate peaks. That part is solid.\n\nWhere it gets soft: the automatic classifier is c = 1 ± 5%, and they admit this may overestimate the floating phase extent. The stress-test note is right that this is not a footnote: the disorder-to-floating transition is KT, so with finite bond dimension a gapped near-critical state can masquerade as c=1 over a sizable window. That means the gray regions in Fig. 1, especially near the KT boundary and the fully coated 1/5-lobe, should be treated as provisional. The 1/5-lobe coating itself is partly extrapolated, not directly simulated. Also, boundaries are guides to the eye with no error bars. None of this kills the main result—there is a genuine floating phase—but it does mean the quantitative extent is not yet nailed down.\n\nCitation pattern looks honest. They cite the relevant prior predictions and related models, and they flag where their results disagree (no floating phase near the 1/2 lobe, contrary to Sela et al.). No code or raw data shipped, which is a shame for a numerical paper, but the method is standard TeNPy and reproducible in principle.\n\nWho should read it: anyone planning Rydberg-chain experiments, and people working on commensurate-incommensurate transitions in 1D. I would bring it to the reading group.\n\nVerdict: send it to peer review. Ask the authors to add a more careful boundary determination—for example, scaling of the structure factor exponent or a KT collapse—or at least show how the c=1 region moves with bond dimension.","headline":"First quantitative map of floating phases in the 1/r^6 Rydberg chain; the central existence claim is solid, but the phase boundaries rest on a c=1 classifier that the authors themselves admit may overreach.","tokens_in":8882,"tokens_out":1727,"would_cite":true,"duration_ms":16673,"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 one-dimensional Rydberg Ising chain with van der Waals interactions hosts extended floating phases, not just direct crystal-to-disorder transitions, and these phases sit at parameters current quantum simulators can reach.","keywords":["Rydberg atoms","floating phase","incommensurate order","Luttinger liquid","quantum phase diagram","van der Waals interaction","tensor network","central charge"],"falsifier":"Measure the static structure factor along the $x=4$ cut of the phase diagram in a Rydberg chain: if sharp Bragg peaks or an energy gap persist where the paper predicts power-law diverging peaks at drifting wavevectors, the floating phase is absent or narrower. Numerically, recomputing the boundaries without the $c=1\\pm 5\\%$ cutoff, for example with explicit Kosterlitz-Thouless scaling fits, would confirm or shrink the reported floating regions.","tokens_in":7806,"feed_emoji":"⚛️","tokens_out":8167,"duration_ms":76507,"temperature":0.7,"pith_summary":"This paper computes the ground-state phase diagram of a one-dimensional chain of Rydberg atoms with van der Waals $1/r^6$ interactions and argues that a third kind of phase, the floating phase, occupies substantial regions of it. In a floating phase the Rydberg density forms a gapless, incommensurate quasi-crystal: correlations decay as a power law and the ordering wavevector drifts continuously with the Rabi frequency and detuning instead of locking to the lattice. The computed floating phases appear between the known $1/q$ crystalline lobes and the disordered phase, fully coat the $1/5$ lobe, and separate the $1/3$ and $1/4$ crystals, at parameters within reach of current Rydberg quantum simulators. The result matters because these phases had been predicted theoretically but not located numerically, and their presence changes how experiments should interpret the melting of the lattice-locked crystals.","feed_headline":"Floating phases fill wide regions of the Rydberg chain phase diagram","feed_subtitle":"These gapless incommensurate solids separate the 1/3 and 1/4 crystals and coat the 1/5 lobe at experimental parameters.","key_machinery":"The central object is the floating phase itself, identified with two numerical diagnostics. The first is the scaling of entanglement entropy $S$ with correlation length $\\xi$, fitted to $S=\\frac{c}{6}\\log\\xi+\\text{const.}$, which yields $c\\approx 1$ in floating regions and $c\\approx 1/2$ at the Ising transition out of the $1/2$ crystal. The second is the static structure factor $S(k)$ computed from density correlations, which shows power-law diverging peaks at a wavevector that drifts incommensurately as parameters are varied and interpolates between the Bragg peaks of neighbouring crystals. The ground states come from infinite matrix product state simulations, with the long-range $1/r^6$ interaction represented as a sum of ten exponentials and gapless energies extrapolated in $1/\\xi^2$.","core_discovery":"The paper's central claim is that the Hamiltonian $H=-\\frac{\\Omega}{2}\\sum_j\\sigma^x_j-\\Delta\\sum_j n_j+V\\sum_{j