{"id":"61814de1-fe2c-4846-8a33-281c69bad6d7","arxiv_id":"2506.19771","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the massive Thirring model with arbitrary Dirichlet boundary fields, the boundary bound states and the SPT/trivial phase structure are solved exactly, and the stability window shrinks with increasing repulsive interactions.","lead":"The authors derive the boundary spectrum of the massive Thirring and sine-Gordon model with generic Dirichlet boundary fields in the repulsive regime, using Bethe ansatz. They find that boundary fields which break charge conjugation destroy boundary zero modes only beyond a critical value, and that this critical value shrinks as bulk interactions strengthen.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The phase boundary φ_C = u rests on the unproven assertion that wide boundary strings never yield localized below-gap states; a numerical Bethe-root census in the uncolored region would settle it.","rationale":"After re-deriving the key steps, the algebraic derivation of the bound-state energies (B73)-(B74) is internally consistent: the residue at ω = iγ in (B47) gives m cosh(γθ), and the corresponding boundary-string integral gives −m cos(γ(u−φ)) = −m sin(γφ) because γu = π/2. The normal-ordering convention does not affect level spacings. Thus the internal arithmetic is not the weak point. The weak point is the claim of completeness: the four states (28) are asserted to be the only states below the two-soliton threshold, and the phase diagram in Fig. 1 is asserted to terminate at |φ| = u. Both assertions rely on the close/wide boundary-string dichotomy inherited from Refs. [40,42,45] and explicitly deferred in Ref. [46]. Since the paper advertises an exact solution, this unverified classification is a genuine gap. The reader's CONDITIONAL verdict is appropriate: the authors should either perform the numerical Bethe-root census described above or provide an analytic argument that wide strings cannot bind states below the gap. My read therefore does not change the reader's verdict.","tokens_in":36662,"tokens_out":10257,"duration_ms":113872,"concrete_test":"Numerically solve the logarithmic form of the Bethe equations (A25) for finite but large system size, e.g., N = 4, L = 10^3, with m0 > 0 and boundary fields just outside region I, e.g., φ_L = u + 0.05, φ_R = u/2, by continuing from a known region-I solution. Enumerate all roots, including complex pairs with Im β ≠ π, and compute each candidate eigenstate's energy via E = m0 Σ cosh β_j. Check whether any complex root yields an energy below the soliton mass m (Eq. 7) and whether the corresponding Bethe wavefunction (16) decays exponentially away from the boundary. Repeating for φ_L = u + δ with δ = 0.01, 0.05, 0.1 would show whether a localized below-gap state persists outside region I; if it does, φ_C = u is not the true phase boundary.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the mid-gap and trivial phases exist only inside regions I and II, with phase boundary at φ_C = u, depends on the classification of boundary Bethe roots into close and wide strings being exhaustive, and on wide strings never producing exponentially localized below-gap states. The paper does not derive this. Section VI states that upon crossing |φ_L,R| = u the close string turns into a wide boundary string and that the localized bound state ceases to exist, while Appendix B states that outside I/II the Bethe equations have wide boundary strings whose analysis is rather involved and will be presented in a further publication. The bound-state energies (30) are computed only for the close-string root ±i(u−φ_b) in Appendix B2c. If, for example, a wide string with rapidity near iπ also produced an eigenstate with |E| < m, then the four-state spectrum (28) would be incomplete and the asserted phase boundaries would not be exact. The authors cite prior work for the close/wide dichotomy, but the uniqueness of the low-energy sector in the presence of two boundaries and arbitrary boundary fields is precisely the new content being claimed, so the inherited classification is load-bearing and is not verified in this setting.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the massive Thirring model (equivalently sine-Gordon) in the repulsive regime with charge-conserving Dirichlet boundary conditions parameterized by boundary fields phi_L and phi_R. Using the coordinate Bethe ansatz, it derives the Bethe equations (18) and claims an exact solution in two boundary-field intervals, region I and region II of Eq. (20). For m0<0 in region I (and by duality for m0>0 in region II) the authors find a trivial phase with a unique ground state; for m0>0 in region I (and by duality for m0<0 in region II) they find a mid-gap phase containing exactly four low-energy states, with boundary bound-state energies m_L,R = m sin(gamma phi_L,R) given by Eq. (30). The paper concludes that the SPT and trivial phases are stable only when the boundary fields lie within a distance phi_C = u of the topological or trivial point, and that this critical width decreases as the bulk interaction strength increases.","tokens_in":36831,"tokens_out":4630,"duration_ms":51318,"significance":"If correct, this is a valuable exact result: it extends the free-fermion boundary phase diagram of the massive Thirring model to all repulsive couplings and demonstrates, within an exactly solvable model, that the stability of SPT edge states under symmetry-breaking boundary fields is controlled by the bulk interaction strength. The manuscript is technically detailed: the Bethe ansatz wave function, integral equations, and renormalized mass/energy calculations are presented explicitly, and the central formulas (21), (30), and (B48) are derived rather than fitted. The main significance, however, is conditional on the completeness of the boundary-root classification, which is the central technical issue raised below.","major_comments":[{"comment":"The phase boundary phi_C = u is load-bearing but is not derived in this manuscript. The text states that outside regions I and II the Bethe equations have wide boundary strings whose analysis 'will be presented in a further publication,' and Section VI asserts that crossing the boundary turns the close string into a wide string and that the localized bound state ceases to exist. The completeness of the close/wide boundary-string classification for the two-boundary problem with arbitrary phi_L,R is inherited from Refs. [40,42] rather than verified here. Since the exact four-state spectrum (28)-(30) and the stability claim rest on there being no other boundary-root configuration with |E| < m, this missing step is load-bearing. I suggest either a proof that wide strings cannot produce imaginary rapidity solutions with |m cosh beta| < m outside regions I/II, or a numerical Bethe-root census over the full phi_L,R domain, which would also clarify the 'uncolored' regions of Fig. 1.","section":"Section VI; Appendix B (paragraph before B.1)"},{"comment":"The abstract's claim of solving the model with 'general Dirichlet boundary conditions' overstates what is shown. The Bethe solution is explicitly restricted to regions I and II of Eq. (20); the uncolored regions of Fig. 1, where wide boundary strings occur, are not solved and are deferred to future work. The abstract and Section VII should be rephrased to state that the solution is obtained within regions I and II, and that the phase diagram outside these regions is not established.","section":"Abstract; Sections I and VII"},{"comment":"The boundary bound-state energies (30) are computed only for the close-string roots +/- i(u - phi_b), and the manuscript does not show that these are the only isolated imaginary solutions of the Bethe equations (A33) in the relevant strip. If an additional isolated imaginary root, for example a wide-string configuration with a rapidity near i pi, also yielded an eigenstate with energy below the mass gap, then the four-state spectrum (28) would be incomplete and the asserted critical value phi_C = u would be incorrect. The paper should either prove the absence of such roots in the two-boundary setting or provide a numerical check for representative values of u and phi_L,R.","section":"Appendix B.2.c; Eq. (B74)"}],"minor_comments":[{"comment":"The figure contains labels B1 and B2, but the caption and Table I only define the A1-A4 sub-phases; the B1/B2 labels should either be defined or removed.","section":"Figure 2"},{"comment":"The sentence stating that for m0<0 the boundary fields are shifted by phi_L,R -> phi_L,R + i pi appears to contain a typo: since phi_L,R are real boundary fields, the shift should be phi_L,R -> phi_L,R + pi.","section":"Appendix B.2, opening paragraph"},{"comment":"There are typos in 'Dirchlet' (should be 'Dirichlet') and 'Luther-Emergy line' (should be 'Luther-Emery line'); these should be corrected.","section":"Appendix A and Section VI"}],"recommendation":"major_revision","confidential_remarks":"The paper leans heavily on the classification of close versus wide boundary strings from Refs. [40,42] by the same group. The editor may wish to ask the authors to state explicitly which parts of that classification are being proved in the present two-boundary setting and which are being assumed, since the completeness of that classification is the crux of the claimed exact phase diagram."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe paper does a real thing: it extends the Bethe ansatz solution of the massive Thirring/sine-Gordon model with Dirichlet boundaries to nonzero boundary fields phi_L,R in the repulsive regime, for two regions of the boundary-field space. The main result is a four-state low-energy spectrum with bound-state energies m_L,R = m sin(gamma phi_L,R), giving a mid-gap phase with sub-phase structure and phase boundaries at |phi_L,R| = u. That is a genuine, non-routine extension of the authors' earlier work [31,32], and the derivation is presented in full: Bethe equations, integral equations, and the energy formula are all there. No fitting parameters; the energies are derived.\n\nThe paper earns credit for being explicit about its own limits. It solves regions I and II only, and it states plainly that outside those regions wide boundary strings appear and are left for future work. That said, the abstract overclaims: 'general Dirichlet boundary conditions' is not what the body delivers. The soft spot is the phase boundary itself. The claim that the mid-gap and trivial phases cease to exist at |phi_L,R| = u rests on the assertion that wide string configurations never produce an exponentially localized below-gap state. The paper does not prove that; it cites earlier work for the close/wide dichotomy, but that dichotomy being exhaustive in exactly this two-boundary, arbitrary-field setting is the new content being claimed. So the phase diagram outside regions I and II is not established. A numerical Bethe-root census or a DMRG check in the uncolored region would settle it.\n\nMy sense: the internal logic of the derivations in regions I and II holds up, and the four-state spectrum is plausible and useful. The weak spot is real but localized; it concerns the boundaries of the phase diagram, not the spectrum inside. The right fix is to soften the abstract and either prove the wide-string no-bound-state claim or add a numerical check.\n\nThis deserves a serious referee. I would send it out. The paper will be useful to people working on superconducting circuits and spin-triplet superconductors where this model is the low-energy description.","headline":"Genuine Bethe-ansatz extension with a derived four-state spectrum; the phase boundary at phi_C = u is load-bearing and unproven, but the paper deserves refereeing.","tokens_in":37396,"tokens_out":1733,"would_cite":true,"duration_ms":18929,"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":"For repulsive interactions, the SPT and trivial phases of the massive Thirring / sine-Gordon model with boundary fields survive only inside a window $|\\phi_{L,R}|<u$, whose width $u$ shrinks as bulk interactions grow, and the edge…","keywords":["massive Thirring model","sine-Gordon model","symmetry protected topological phase","Bethe ansatz","boundary bound states","duality symmetry","charge conjugation symmetry","mid-gap states"],"falsifier":"Solve the Bethe equations in the uncolored regions $|\\phi_{L,R}|>u$ for a wide boundary string whose energy $E=m_0\\sum_j\\cosh\\beta_j$ lies below the soliton mass $m$; alternatively, numerically diagonalize the lattice sine-Gordon model with Dirichlet boundary fields and look for a localized state with energy below $m$ at $|\\phi|>u$. Either observation would refute the claimed phase boundary at $\\phi_C=u$.","tokens_in":36420,"feed_emoji":"⚛️","tokens_out":11660,"duration_ms":110045,"temperature":0.7,"pith_summary":"The paper solves the one-dimensional massive Thirring model (equivalently, the sine-Gordon model) in the repulsive regime with general Dirichlet boundary conditions, described by left and right boundary fields $\\phi_L,\\phi_R$ that explicitly break charge conjugation. It establishes that the trivial and symmetry-protected topological (SPT) phases remain stable away from their special symmetry points only while the boundary fields stay within a window of half-width $u$ around the trivial or topological point; outside that window the low-energy description changes qualitatively. Inside the window the boundary spectrum is exact and simple: two edge bound states with energies $m_{L,R}=m\\sin(\\gamma\\phi_{L,R})$, giving four low-energy states in the mid-gap phase and a unique ground state in the trivial phase. If correct, this fixes the boundary phase diagram for all repulsive couplings and shows that the stability of topological protection is controlled by the interplay of boundary-field values and bulk interaction strength.","feed_headline":"Stronger interactions shrink the SPT phase window to zero","feed_subtitle":"Exact solution shows edge bound states vanish exactly at the critical field u set by bulk interactions.","key_machinery":"The argument is carried by the coordinate Bethe ansatz for the massive Thirring model with general open boundary conditions. The Bethe equations include boundary-dependent factors with the anomalous shift $u-\\pi/2$, and the key objects are close (short) boundary strings: purely imaginary rapidity solutions that describe bound states exponentially localized at an edge. Adding or removing such a string from the reference states $|\\pm1\\rangle$ generates the four-state boundary Hilbert space, and its energy evaluates to $\\mp m\\sin(\\gamma\\phi_b)$; the phase boundary at $|\\phi_b|=u$ is where a close string turns into a wide boundary string, which no longer describes a simple localized bound state. A duality symmetry $\\Omega$ (changing the sign of $m_0$ and shifting $\\phi_{L,R}\\to\\phi_{L,R}+\\pi$) maps the trivial and mid-gap regions onto each other and lets the authors solve both phases from one calculation.","core_discovery":"The central claim is that the massive Thirring model with open boundaries and generic boundary fields has, in the repulsive regime, two exact phases that survive away from their symmetry points. At the topological point $m_0>0$, $\\phi_L=\\phi_R=0$, and at its duality image $m_0<0$, $\\phi_L=\\phi_R=\\pi$, the system hosts zero-energy bound states at both edges and a twofold degenerate ground state in each fermionic parity sector; this is the SPT phase. For boundary fields inside region I, $-u<\\phi_{L,R}<u$ (for $m_0>0$), the degeneracy is lifted but the boundary Hilbert space is unchanged: the four states $|-1\\rangle$, $|0\\rangle_L$, $|0\\rangle_R$, $|+1\\rangle$ have energies $E_{|0\\rangle_{L,R}}=E_{|-1\\rangle}+m_{L,R}$ and $E_{|+1\\rangle}=E_{|-1\\rangle}+m_L+m_R$ with $m_{L,R}=m\\sin(\\gamma\\phi_{L,R})$, $\\gamma=\\pi/(2u)$. As a boundary field approaches $|\\phi|=u$ the corresponding bound state rises to the soliton mass and merges with the continuum; beyond $u$ the close boundary string turns into a wide boundary string and both the mid-gap and trivial phases cease to exist. By duality, $m_0\\to-m_0$ and $\\phi_{L,R}\\to\\phi_{L,R}+\\pi$ interchanges the two regions, so the same statement holds for $m_0<0$ with region II playing the role of the mid-gap phase. The critical half-width $u$ is largest at the Luther-Emery (free-fermion) point $\\beta=\\sqrt{4\\pi}$ and decreases to zero as the interaction strength increases toward $\\beta\\to\\sqrt{8\\pi}$.","pith_inferences":["A natural next test is a numerical diagonalization or tensor-network simulation of the lattice sine-Gordon model with Dirichlet walls: the paper predicts edge states at exactly $m\\sin(\\gamma\\phi)$ and their disappearance at $|\\phi|=u$, and a localized state surviving past $u$ would mean the string classification is incomplete.","The paper leaves the region outside I and II to a wide-boundary-string regime; if the close-to-wide transition is a genuine boundary quantum phase transition, the exact value $u$ gives a sharp location to look for nonanalytic changes in edge entanglement or local density.","In the attractive regime the paper states that boundary breathers appear and the spectrum is richer; by analogy, boundary fields there are likely to split or shift breather levels, so the four-state structure is probably specific to repulsive interactions.","The interaction dependence of $u$ suggests an experimental route: by tuning bulk interactions one can continuously vary how much boundary asymmetry an SPT phase tolerates, turning edge stability itself into a probe of the bulk Luttinger parameter."],"forward_implications":["Inside the stability window, the SPT ground-state degeneracy is replaced by a fixed ladder of three mid-gap states, so the boundary Hilbert space has dimension four for every $\\phi_{L,R}$ in region I or II.","The bound-state energies $m\\sin(\\gamma\\phi_{L,R})$ are continuously tunable between $-m$ and $m$; changing the signs of $\\phi_L,\\phi_R$ selects which of the four states is the ground state (subphases $A_1$ through $A_4$).","Because the critical value $\\phi_C=u$ shrinks as repulsion strengthens, the SPT phase is maximally tolerant of boundary-field asymmetry at the Luther-Emery free-fermion point and loses all tolerance as $\\beta\\to\\sqrt{8\\pi}$.","The duality symmetry makes the trivial phase for $m_0<0$, $\\phi_{L,R}\\in(-u,u)$ exactly equivalent to the mid-gap phase for $m_0>0$, $\\phi_{L,R}$ in region II, so results for one mass sign transfer quantitatively to the other.","Any physical realization of the sine-Gordon or massive Thirring model with tunable boundary fields has a precise prediction for where edge bound states appear and at what energy."],"supporting_citations":[{"why":"Supplies the bulk Bethe-ansatz solution of the massive Thirring model, the soliton mass formula, and the renormalized charge used throughout.","marker":"[25]"},{"why":"Provides the open-boundary Bethe-ansatz framework and the boundary anomaly $u-\\pi/2$ that fixes the relation to sine-Gordon boundary fields.","marker":"[38]"},{"why":"Exact solution of the U(1) Thirring model showing zero-energy modes survive weak coupling and introducing close boundary strings.","marker":"[22]"},{"why":"Recent exact solution of the sine-Gordon/massive Thirring model establishing the SPT phase at special boundary fields, which this paper extends to generic fields.","marker":"[31]"},{"why":"Weak-coupling analysis of the duality between trivial and topological boundary fixed points that the present duality argument generalizes.","marker":"[32]"},{"why":"Classifies close versus wide boundary strings, the basis for asserting which Bethe roots represent exponentially localized edge states.","marker":"[40]"},{"why":"Further develops the close and wide boundary string classification used to exclude localized states outside regions I and II.","marker":"[42]"},{"why":"Describes wide boundary strings, used to characterize the regime outside the stability windows where the simple phases cease to exist.","marker":"[45]"}],"fun_headline_variants":["Critical boundary field shrinks with interaction strength","Interaction strength sets SPT stability boundary","SPT phase window closes as interactions increase","SPT survival depends on boundary field and bulk interaction","Stronger interactions close SPT phase window"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the Bethe-ansatz classification into close and wide boundary strings is complete, so that inside regions I and II no other exponentially localized state exists below the soliton mass; if a wide-string configuration produced an additional localized bound state, the critical value $\\phi_C=u$ and the phase boundaries would be wrong.","fun_headline_variants_meta":{"raw":{"variants":["Critical boundary field shrinks with interaction strength","Interaction strength sets SPT stability boundary","SPT phase window closes as interactions increase","SPT survival depends on boundary field and bulk interaction","Stronger interactions close SPT phase window"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000761,"raw_usage":{"total_tokens":3545,"prompt_tokens":1279,"completion_tokens":2266,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":895,"completion_tokens_details":{"reasoning_tokens":2199}},"tokens_in":895,"tokens_out":2266,"duration_ms":16680,"temperature":1.0,"reasoning_tokens":2199,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:24:48.256476+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the Bethe equations in the uncolored regions $|\\phi_{L,R}|>u$ for a wide boundary string whose energy $E=m_0\\sum_j\\cosh\\beta_j$ lies below the soliton mass $m$; alternatively, numerically diagonalize the lattice sine-Gordon model with Dirichlet boundary fields and look for a localized state with energy below $m$ at $|\\phi|>u$. Either observation would refute the claimed phase boundary at $\\phi_C=u$.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the bulk Bethe-ansatz solution of the massive Thirring model, the soliton mass formula, and the renormalized charge used throughout."},{"cited_title":"Seiler and D","cited_arxiv_id":null,"evidence_quote":"Provides the open-boundary Bethe-ansatz framework and the boundary anomaly $u-\\pi/2$ that fixes the relation to sine-Gordon boundary fields."},{"cited_title":"Japaridze, A","cited_arxiv_id":null,"evidence_quote":"Classifies close versus wide boundary strings, the basis for asserting which Bethe roots represent exponentially localized edge states."}],"review_version":2}