{"id":"42b146d1-510a-4c75-b750-82279f46e77f","arxiv_id":"1908.08172","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The boundary energy of the Lieb-Liniger Bose gas in a hard-wall box is fixed by an exact integral equation, with new weak- and strong-coupling expansions that differ from the dark-soliton energy.","lead":"The authors derive an exact integral equation for the boundary energy of one-dimensional repulsive bosons in a hard-wall box, correcting a 1971 approximation by Gaudin. The result matters for cold-atom experiments in flat box traps, where surface finite-size effects become measurable.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exactness claim rests on an unproven discrete-to-continuum limit between Eqs. (11) and (14); no bound on the O(1/N) remainder is given, so the exact integral equation is a conjecture unless verified.","rationale":"The paper's main result is a clean integral equation for the boundary energy of the Lieb-Liniger model, with plausible asymptotic expansions, a correction to Gaudin's earlier result, and independent checks from Gross-Pitaevskii theory, dual fermion perturbation theory, and quantum Monte Carlo. The strongest mathematical claim, however, is that Eq. (14) is exact in the thermodynamic limit for arbitrary c>0. That claim rests on replacing finite-N sums by integrals weighted by the periodic Lieb density and on linearizing in Delta k_i, with a remainder written only as O(1/N). No rigorous bound is supplied, and the behavior of the remainder near the Fermi edges at weak coupling is delicate because the density vanishes there. This is precisely the weakest assumption identified by the reader. It does not make me think the result is wrong; the numerics and cross-checks are reassuring. But it does mean the paper is conditional: the exactness claim would be fully established by a controlled estimate of the remainder or by a direct numerical demonstration that the finite-N boundary energy converges to the integral-equation value at each fixed gamma. Since the reader explicitly conditioned the verdict on this same issue, I recommend no change to the verdict.","tokens_in":12359,"tokens_out":26028,"duration_ms":254761,"concrete_test":"For fixed gamma (e.g., 0.001, 0.01, 0.1, 1, 10, 100), solve the discrete Bethe equations (6) and (8) for N=50, 100, 200, 400, 800 and compute E_B^{(N)} = (hbar^2/2m) Σ (kbar_i^2 - k_i^2). Solve Eqs. (12), (14), and (15) numerically to obtain E_B^{(∞)}. Check whether |E_B^{(N)} - E_B^{(∞)}| → 0 as N→∞, for example by fitting N^α|ΔE| and verifying α>0 and N|ΔE| → 0. Record how α and the prefactor vary with gamma. If the difference does not vanish for some fixed gamma, the continuum-limit step is invalid; if it vanishes but the prefactor grows like γ^{-β}, the weak-coupling asymptotic expansion (19) requires an additional argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that Eq. (14) with Eq. (15) gives the exact thermodynamic-limit boundary energy for every c>0 depends on the step between Eqs. (11) and (14), where discrete Bethe-ansatz sums are replaced by integrals against the continuous Lieb density rho(k). Equation (13) is asserted in the thermodynamic limit but is then used to eliminate the finite-N sum in Eq. (11); the paper provides no estimate of the difference between the discrete sum and 2L∫rho dk, nor of the Taylor remainder in Delta k_i. The O(1/N) symbol in Eq. (11) is stated, not bounded. The concern is not only cosmetic: near the Fermi edges the Lieb density vanishes as a square root at weak coupling, so the local quasimomentum spacing can be as large as N^{-2/3}, making the continuum replacement slower than the usual 1/N. If the remainder fails to vanish as N→∞ for some c>0, Eq. (14) is not exact; if it vanishes only nonuniformly in c, the weak-coupling expansion (19) needs separate justification. The text acknowledges finite-size corrections only in passing (last paragraph) and does not address this exchange of limits.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the ground-state boundary energy EB of one-dimensional bosons with contact interaction (Lieb-Liniger model) in a hard-wall box. Starting from Gaudin's Bethe-Ansatz equations for zero and periodic boundary conditions, the authors subtract the two sets of equations and, after a discrete-to-continuum step, obtain a linear integral equation (14) for an odd function g(k) that determines EB through Eq. (15). This is claimed to be the exact thermodynamic-limit boundary energy for arbitrary interaction c>0. The paper then solves the integral equation in the weakly and strongly interacting limits, obtaining EB=(8/3)ε√γ[1−3√γ/16+O(γ)] and EB=(π²/2)ε[1−4/(3γ)−4/γ²+O(γ⁻³)], and compares the results with quantum Monte Carlo data and with a dual-fermion perturbation calculation. It also shows that Gaudin's earlier boundary-energy expression coincides with the energy of the zero-velocity type-II excitation and overestimates the true boundary energy, and it argues for a universal leading form of EB in terms of the gas parameter using the hard-rod gas as a comparison.","tokens_in":12585,"tokens_out":11764,"duration_ms":119722,"significance":"If the central derivation is accepted, the paper settles a long-standing question: it provides a compact integral-equation characterization of the boundary energy of the Lieb-Liniger model, corrects Gaudin's 1971 result beyond leading order, and exposes the distinction between the boundary energy and the dark-soliton/type-II-excitation energy. The result is elegant and should be useful for finite-size analyses of one-dimensional Bose gases in box potentials and for impurity problems with an infinite repulsive wall. The manuscript is strengthened by independent numerical checks (diffusion Monte Carlo) and by an analytic dual-fermion check of the strong-coupling expansion. The derivations are algebraic and the integral equations used are standard for the Lieb-Liniger model, and no free parameters are fitted to the boundary energy.","major_comments":[{"comment":"The central exactness claim is not fully established because the transition from the finite-N Bethe-Ansatz equations to the integral equation (14) replaces sums over the N positive quasimomenta by integrals against the continuum density 2Lρ(k) with no estimate of the remainder. Equation (11) carries an O(1/N) symbol, but the error in approximating Σ_j [θ′(k_i−k_j)−θ′(k_i+k_j)]Δk_j by the corresponding integral is not bounded. This is load-bearing rather than cosmetic because EB is an O(1) quantity obtained from an O(1/N) difference of two large energies, so an uncontrolled O(1/N) remainder could in principle contribute. At weak coupling the Lieb density vanishes as (Q−k)^{1/2} near the Fermi edge, so the local quasimomentum spacing there is much larger than 1/N and standard Euler-Maclaurin-style estimates fail; the limit N→∞ may also be nonuniform in γ. To support the statement that Eq. (14) is exact for all c>0, the authors should either supply a justified remainder estimate (or cite a known theorem for this continuum limit), or explicitly restate the status of Eq. (14) as an assumption/conjecture rather than a proven exact result. The same issue affects the weak-coupling expansion (19) if the interchange of N→∞ and γ→0 is not controlled.","section":"Equations (11)-(14)"},{"comment":"The weak-coupling solution ρ(x) and σ(x) used to derive Eq. (19) is stated to be valid only for 1−x²≫λ, yet the derivation integrates these expressions over the entire interval including the Fermi-edge region. The text asserts that this limitation is not important, but gives no estimate of the boundary-layer contribution. Because Eq. (19) contains a genuine subleading correction (the −3√γ/16 term), the authors should show explicitly that the region 1−x²≲λ contributes only at higher order in γ to λ, σ, and ultimately EB; otherwise the quoted first correction is not justified by the presented calculation.","section":"Supplemental Material, Eqs. (S6)-(S9)"}],"minor_comments":[{"comment":"The two terms in the square bracket, θ′(k_i−k_j)(Δk_i−Δk_j) and θ′(k_i+k_j)(Δk_i+Δk_j), appear without a visible plus sign between them; please correct the typography.","section":"Equation (11)"},{"comment":"The name in the acknowledgment is typeset as \"V . Yurovky\"; it should read \"V. A. Yurovsky\".","section":"Acknowledgment"},{"comment":"The statement that the N=21 and N=41 quantum Monte Carlo results agree with the Bethe-Ansatz curve would be more persuasive with a quantitative statement of the finite-N difference and, if possible, an extrapolation to the thermodynamic limit; the current text relies on visual inspection of the figure.","section":"Figure 1 and QMC paragraph"},{"comment":"The notation γ=−2/na in Eq. (22) reuses the symbol γ, which was earlier defined as c/n>0. Please clarify that this is a different, negative gas parameter for the hard-rod mapping, so that the comparison with Eq. (20) is unambiguous.","section":"Hard-rod universality section"},{"comment":"The sentence about the regime L≲ξ would benefit from a more explicit statement of the order of limits: the derivation of Eq. (15) takes L→∞ with density fixed, and the very-weak-coupling finite-size regime γ≲1/N² is a separate limit that the paper deliberately does not address.","section":"Last paragraph"}],"recommendation":"major_revision","confidential_remarks":"The main technical uncertainty is the discrete-to-continuum limit between Eqs. (11) and (14). This is a common step in Bethe-Ansatz derivations, and the algebraic content of the paper checks out, but the word \"exact\" in the title and text is stronger than what is proved. If the authors can supply a justified remainder estimate, cite a rigorous treatment, or explicitly downgrade the claim to a standard but unproven continuum-limit assumption, I would be willing to support publication. The weak-coupling boundary-layer issue in the Supplemental Material should also be addressed for the quoted subleading term."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something concrete and useful. It shows that Gaudin's 1971 boundary-energy expression for the Lieb-Liniger gas is actually the energy of the type-II excitation, and it derives the correct integral equation for the true boundary energy, Eq. (14), with the corrected right-hand side r(k)=sgn(k)/4 + (1/2π) arctan(2k/c). That correction matters: Gaudin's expression overestimates the boundary energy by a factor of two in the Tonks limit and differs at subleading order in weak coupling. The new weak- and strong-coupling expansions, Eqs. (19) and (20), look right, and the agreement with quantum Monte Carlo and with the dual-fermion perturbation theory in the supplement is the kind of independent check I trust.\n\nWhat is new: the integral equation (14), the subleading expansions, and the identification of Gaudin's r_G with the type-II excitation at all γ. That identification is clearly stated and, as far as I can tell, correct.\n\nNow the soft spot. The step from the discrete Bethe-ansatz equations (11) to the continuous integral equation (14) replaces sums by integrals weighted by the Lieb density, and the paper simply writes O(1/N) without a bound. The stress-test note is right that near the Fermi edges the quasimomentum spacing is not uniformly 1/N, especially at weak coupling, so the remainder could be nonuniform in γ. This is a real gap in the presentation, but I do not think it is a load-bearing flaw. The same continuum-limit machinery is standard for the Lieb equation, the final equation reduces correctly in both limits, and the numerics—both the discrete Bethe-ansatz solution and QMC—track the predicted curve. If the exactness claim rests on a conjecture, it is a well-tested conjecture. For a Letter, I would ask the authors to either discuss the remainder more carefully or soften \"exact\" to \"thermodynamic-limit\" with a caveat, but I would not reject over it.\n\nMinor issues: the QMC points in Fig. 1 have no visible error bars, and the hard-rod universality comparison uses γ=-2/na without spelling out the mapping. Both are easy fixes.\n\nWho is this for: anyone working on 1D Bose gases, boundary effects, or integrable models. It deserves a serious referee and, after minor revision, publication.","headline":"A genuinely useful correction of Gaudin's boundary-energy formula, with a new exact integral equation; the one real caveat is a terse continuum-limit step, but the result is well supported and deserves peer review.","tokens_in":13168,"tokens_out":4993,"would_cite":true,"duration_ms":45897,"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":"Exact boundary energy of a 1D Bose gas is fixed by a corrected integral equation","keywords":["Lieb-Liniger model","boundary energy","Bethe ansatz","hard-wall box","integral equation","dark soliton","one-dimensional bosons","quantum Monte Carlo"],"falsifier":"Solve the Bethe-ansatz equations (6) and (8) numerically for $N = 100, 1000, 10000$ at several interaction strengths (e.g., $\\gamma = 0.1, 1, 10$), compute $\\Delta k_i$, and form $g_N(k) = L\\rho(k)\\Delta k_i$; if $g_N(k)$ does not approach the solution of Eq. (14) with corrections that vanish as $1/N$ uniformly in $\\gamma$, then the integral equation is not exact. A finite-$N$ extrapolation of $E_B$ from the Bethe equations that disagrees with Eq. (15) at any $\\gamma$ would also settle it.","tokens_in":12154,"feed_emoji":"⚛️","tokens_out":5669,"duration_ms":50645,"temperature":0.7,"pith_summary":"This paper claims to settle the boundary energy of the one-dimensional Lieb-Liniger Bose gas confined by hard walls: the energy cost of the walls' presence is fixed exactly, for any strength of inter-particle repulsion, by a single linear integral equation. Gaudin's 1971 Bethe-ansatz treatment contained an equation of the same form but with the wrong right-hand side; it actually computes the energy of a moving dark soliton, not the boundary energy. With the corrected right-hand side the authors obtain the weak-coupling expansion $E_B = \\frac{8}{3}\\epsilon\\sqrt{\\gamma}\\,[1 - \\frac{3}{16}\\sqrt{\\gamma} + O(\\gamma)]$ and the strong-coupling expansion $E_B = \\frac{\\pi^2}{2}\\epsilon\\,[1 - \\frac{4}{3\\gamma} - \\frac{4}{\\gamma^2} + O(\\gamma^{-3})]$, and verify the result against diffusion Monte Carlo simulations. Because the boundary energy is the leading finite-size correction in box-trapped atomic gases, an exact formula is a direct check on experiments and on approximate treatments of inhomogeneous one-dimensional quantum gases.","feed_headline":"Exact boundary energy of 1D Bose gas found for all interactions","feed_subtitle":"Corrects a 1971 formula and separates boundary energy from dark-soliton energy.","key_machinery":"The load-bearing object is the linear integral equation for the odd function $g(k) = L\\rho(k)(\\bar{k}_i - k_i)$, obtained by subtracting the Bethe ansatz equations for zero and periodic boundary conditions and replacing the discrete quasimomentum sums by integrals over the Lieb density $\\rho(k)$. The correct source term $r(k) = \\operatorname{sgn}(k)/4 + \\frac{1}{2\\pi}\\arctan\\frac{2k}{c}$ carries the boundary condition's effect; it is the only place where Gaudin's equation is altered. The paper then reformulates the problem through the symmetric Green's function of the Lieb equation, defining $\\sigma(k)$ that satisfies the same integral equation with source $k$, so that $E_B = \\int \\sigma(k) r(k)\\, dk$. This Green's-function form is what makes the weak- and strong-coupling asymptotic expansions tractable.","core_discovery":"The central result is the integral equation $g(k) - \\frac{c}{\\pi}\\int_{-Q}^{Q} dk'\\, \\frac{g(k')}{c^2 + (k'-k)^2} = r(k)$ with the source term $r(k) = \\frac{\\operatorname{sgn}(k)}{4} + \\frac{1}{2\\pi}\\arctan\\frac{2k}{c}$, together with the energy formula $E_B = \\frac{\\hbar^2}{m}\\int_{-Q}^{Q} k\\, g(k)\\, dk$. The paper argues that this pair gives the exact thermodynamic-limit boundary energy for every repulsion strength $c > 0$, because the difference between the zero- and periodic-boundary Bethe ansatz quasimomenta becomes, in the thermodynamic limit, a smooth odd function $g(k)$ solving this equation. The correction to Gaudin is the extra $\\frac{1}{2\\pi}\\arctan\\frac{2k}{c}$ term: Gaudin's source $r_G(k) = \\operatorname{sgn}(k)/2$ is only the $c\\to 0$ limit. Consequently, what Gaudin computed is exactly the energy of Lieb's type-II excitation at momentum $\\pi\\hbar n$ (the dark soliton in the weak-coupling limit), which is always larger than the true boundary energy. The paper also shows that the first two terms of the strong-coupling expansion are universal, matching the exact hard-rod gas result when written in terms of the gas parameter.","pith_inferences":["If the equation is exact, finite-$N$ corrections to the boundary energy could be computed systematically by expanding around the thermodynamic-limit $g(k)$, giving a route to quantitative finite-size scaling in one-dimensional traps.","The distinction between the type-II excitation and the boundary energy suggests that other 'boundary' quantities defined through Bethe ansatz comparisons may similarly hide a soliton energy; the same subtract-and-expand method could be applied to spin chains or Fermi gases in hard-wall boxes.","A testable extension: measure the density profile near a hard wall in an ultracold one-dimensional Bose gas; the boundary energy's dependence on $\\gamma$ implies a specific healing-length correction that time-of-flight imaging could resolve."],"forward_implications":["The exact boundary energy can be evaluated numerically for any $\\gamma$, providing a benchmark for finite-size corrections in box-trapped one-dimensional Bose gases.","The subleading term in the weak-coupling boundary energy gives a precise beyond-mean-field prediction that can be compared with experiments and with Gross-Pitaevskii-based treatments.","Gaudin's dark-soliton identification is corrected: the dark soliton and the boundary energy are distinct except at leading weak-coupling order, settling a long-standing confusion.","The universality of the first two strong-coupling coefficients extends to hard-rod gases and other short-range-interacting one-dimensional systems, so the boundary energy depends only on the gas parameter at low density.","The asymptotic expansions together with the integral equation supply a controlled interpolation across the crossover from Tonks-Girardeau to mean-field regimes."],"supporting_citations":[{"why":"Supplies the original zero-boundary Bethe ansatz and the integral equation for the boundary energy that this paper corrects.","marker":"[22]"},{"why":"Provides the Lieb-Liniger model and the integral equation for the quasimomentum density $\\rho(k)$ used to take the continuum limit.","marker":"[8]"},{"why":"Defines the type-II excitations whose energy is shown to coincide with Gaudin's boundary-energy formula.","marker":"[19]"},{"why":"Provides the fermion mapping used for the Tonks-Girardeau limit and the hard-rods comparison.","marker":"[29]"},{"why":"Supplies the weak-coupling solution of the Lieb density equation used to derive the expansion (19).","marker":"[33]"},{"why":"Supplies the strong-coupling iterative solution of the integral equations used for expansion (20).","marker":"[35]"},{"why":"Provides the dual Cheon-Shigehara fermion model used as an independent check of the strong-coupling correction.","marker":"[30]"}],"fun_headline_variants":["Exact boundary energy for 1D bosons at any interaction","New integral equation nails 1D boson boundary energy","Correcting Gaudin: exact 1D boson boundary energy","Separating boundary energy from dark soliton in 1D"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation replaces the discrete Bethe-ansatz equations by a continuum integral equation, assuming that the quasimomentum shifts $\\Delta k_i$ are $O(1/L)$ and converge to a smooth odd function $g(k)$ with a remainder that vanishes uniformly as $N\\to\\infty$ for every interaction strength $c>0$; if that remainder is not uniform in $c$, the claimed exactness at arbitrary interaction would be lost.","fun_headline_variants_meta":{"raw":{"variants":["Exact boundary energy for 1D bosons at any interaction","New integral equation nails 1D boson boundary energy","Correcting Gaudin: exact 1D boson boundary energy","Separating boundary energy from dark soliton in 1D"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000521,"raw_usage":{"total_tokens":2557,"prompt_tokens":1014,"completion_tokens":1543,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":630,"completion_tokens_details":{"reasoning_tokens":1471}},"tokens_in":630,"tokens_out":1543,"duration_ms":10642,"temperature":1.0,"reasoning_tokens":1471,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:48:25.878374+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the Bethe-ansatz equations (6) and (8) numerically for $N = 100, 1000, 10000$ at several interaction strengths (e.g., $\\gamma = 0.1, 1, 10$), compute $\\Delta k_i$, and form $g_N(k) = L\\rho(k)\\Delta k_i$; if $g_N(k)$ does not approach the solution of Eq. (14) with corrections that vanish as $1/N$ uniformly in $\\gamma$, then the integral equation is not exact. A finite-$N$ extrapolation of $E_B$ from the Bethe equations that disagrees with Eq. (15) at any $\\gamma$ would also settle it.","supporting_citations":[{"cited_title":"Boundary Energy of a Bose Gas in One Dimen- sion,","cited_arxiv_id":null,"evidence_quote":"Supplies the original zero-boundary Bethe ansatz and the integral equation for the boundary energy that this paper corrects."},{"cited_title":"Exact Analysis of an In- teracting Bose Gas. I. The General Solution and the Ground State,","cited_arxiv_id":null,"evidence_quote":"Provides the Lieb-Liniger model and the integral equation for the quasimomentum density $\\rho(k)$ used to take the continuum limit."},{"cited_title":"Exact Analysis of an Interacting Bose Gas. II. The Excitation Spectrum,","cited_arxiv_id":null,"evidence_quote":"Defines the type-II excitations whose energy is shown to coincide with Gaudin's boundary-energy formula."},{"cited_title":"Relationship between Systems of Impenetrable Bosons and Fermions in One Dimension,","cited_arxiv_id":null,"evidence_quote":"Provides the fermion mapping used for the Tonks-Girardeau limit and the hard-rods comparison."},{"cited_title":"Theory of one-dimensional Bose gas with point interaction,","cited_arxiv_id":null,"evidence_quote":"Supplies the weak-coupling solution of the Lieb density equation used to derive the expansion (19)."},{"cited_title":"Excitation Spectrum of the Lieb-Liniger Model,","cited_arxiv_id":null,"evidence_quote":"Supplies the strong-coupling iterative solution of the integral equations used for expansion (20)."},{"cited_title":"Fermi-Bose mapping for 6 one-dimensional Bose gases,","cited_arxiv_id":null,"evidence_quote":"Provides the dual Cheon-Shigehara fermion model used as an independent check of the strong-coupling correction."}],"review_version":1}