{"id":"cd2ff4b9-3a27-4f8b-9239-0015f7abb5a2","arxiv_id":"1908.02100","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In a two-field BPS sine-Gordon model, static soliton solutions are numerically shown to exhibit very slow attraction or repulsion caused by numerical noise exciting extra zero modes.","lead":"This paper numerically studies collisions and bound states of kinks and antikinks in two coupled sine-Gordon-like fields. It reports that supposedly static multi-field BPS solutions can drift slowly because tiny numerical errors excite additional zero modes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero-mode explanation is unsupported: no projection of numerical error onto the internal zero mode is measured, and energy conservation is blind to motion along the flat direction.","rationale":"The paper is a legitimate numerical exploration, and the zero-mode mechanism is physically plausible: for a two-field BPS system there is an internal modulus in addition to translation, so numerical noise can excite a slow drift that conserves energy to high accuracy. However, the central claim is presented without the quantitative evidence needed to distinguish zero-mode excitation from discretization artifacts or radiation from the sewn initial data. The reader's weakest_assumption already identifies exactly this gap: only global energy conservation is reported, with no convergence study and no mode-amplitude measurement. My proposed test would settle the mechanism directly by projecting the initial error onto the zero modes and removing that component. If the drift disappears, the claim stands; if it persists, the paper's main conclusion is not supported. Because the reader's conditional verdict already demands such confirmation, my analysis does not change the verdict; it sharpens the condition under which the paper should be accepted.","tokens_in":19616,"tokens_out":8314,"duration_ms":102795,"concrete_test":"Compute the linearized zero modes around the numerical BPS solution for lambda = +/-1.8 kink-antikink and lambda = +/-1.95 kink-kink by solving the Jacobi operator eigenproblem; project the initial-data error (the difference from a high-resolution reference solution) onto the translation and internal zero modes. Then rerun the evolution with the internal zero-mode component subtracted. If the slow drift vanishes, the zero-mode explanation is confirmed; if it persists, the drift is caused by discretization or radiation and the central claim fails. As a complementary check, repeat at spatial spacings dx = 0.1, 0.05, 0.025 and in double versus quadruple precision to verify that the drift amplitude and sign track the numerical error amplitude rather than remaining fixed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 5.3 is that the slow drift of ostensibly static BPS solutions is caused by excitation of the second zero mode by tiny numerical errors. This requires that (i) the initial numerical field differs from the exact BPS solution mainly along the zero-mode direction, and (ii) the observed drift is not instead a truncation or radiation artifact of the sewn initial data. Neither is established. The only global diagnostic reported is energy conservation to 10^-5% (Figs. 26, 37, 38), but the zero mode is precisely the flat direction of the energy functional, so energy conservation cannot constrain motion along it. No convergence study in dx/dt is given, no single- versus double-precision comparison, and no projection of the initial error onto the two zero modes is performed. The systematic sign dependence (attraction/repulsion governed by sign(lambda) and kink/antikink sector, Figs. 33-38) is also hard to attribute to random roundoff; deterministic discretization error could produce the same pattern, which would make the phenomenon a numerical artifact of the scheme rather than generic zero-mode excitation. The paper itself concedes that the sign dependence is unexplained at the end of Section 5.4.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a (1+1)-dimensional model of two coupled sine-Gordon-like scalar fields with a BPS sector, continuing earlier work on the same system. Using numerically determined one-soliton BPS solutions as building blocks, the authors construct two-soliton and kink-antikink initial configurations, evolve them in the Lorentz-covariant field equations, and catalogue their interactions as a function of the coupling λ. They report that two-kink systems repel for small λ but form bound states and emit breathers for larger λ, while kink-antikink systems show sign-dependent attraction/repulsion and oscillon formation. The central new claim, developed in Section 5, is that ostensibly static BPS solutions of the multi-field model slowly drift because tiny numerical errors excite one of the system's zero modes; the direction of the drift is reported to depend on the sign of λ and on the kink/antikink sector. The paper is almost entirely numerical, with no derivation of the drift rate, no convergence study, and no specification of the numerical scheme.","tokens_in":19879,"tokens_out":17533,"duration_ms":180480,"significance":"If the central claim is correct, the paper identifies a practically important phenomenon: long-time numerical simulations of multi-field BPS solitons can exhibit slow, spurious motion along zero-mode directions, which could contaminate studies of soliton interactions. The authors deserve credit for running very long simulations, for reporting energy conservation at the 10^-5% level in several runs, and for clearly separating kink-kink from kink-antikink sectors. The observation that the sign of the drift depends systematically on λ and on the sector is an interesting empirical pattern. However, the paper provides no direct evidence that numerical error projects onto the claimed zero mode, and the authors themselves state at the end of Section 5.4 that they have not explained the sign dependence. Because the entire conclusion rests on the interpretation of numerical observations, the absence of convergence tests, mode-amplitude diagnostics, and a description of the discretization is a serious gap. The result is potentially useful as a cautionary note to the soliton community, but it is not yet established.","major_comments":[{"comment":"The central claim that the observed slow drift is caused by excitation of the second zero mode by numerical errors is not supported by any direct measurement. No projection of the initial numerical error, or of the evolving field, onto the two zero modes is reported; no convergence study in dx or dt is given; and no single- versus double-precision comparison is made. Energy conservation to 10^-5% (Figs. 26, 37, 38) cannot constrain motion along a zero mode, because the energy is flat along that direction by definition. In addition, the numerical scheme (discretization, time integrator, grid spacing, time step, boundary treatment) is not described anywhere in the paper, so the reported drifts cannot be checked or reproduced. A refinement study and a mode-projection diagnostic are needed to distinguish genuine zero-mode excitation from a discretization artifact.","section":"§5.3"},{"comment":"The systematic sign dependence reported in Figs. 33–38 (attraction for λ<0 in the kink-antikink sector and for λ>0 in the kink-kink sector, with repulsion in the opposite cases) is difficult to attribute to 'small unavoidable numerical errors,' which one would expect to be essentially random. Deterministic discretization error could produce exactly such systematic behavior, which would make the phenomenon an artifact of the numerical scheme rather than generic zero-mode excitation. The authors explicitly concede at the end of Section 5.4 that they have not understood why the sign of λ selects attraction or repulsion; this gap is load-bearing because a zero-mode mechanism should predict the drift direction from the structure of the perturbation. At minimum, the sign of the zero-mode component of the numerical error should be measured and shown to correlate with the observed drift direction.","section":"§5.4"},{"comment":"The paper does not clearly distinguish initial conditions that are exact multi-soliton BPS solutions from configurations obtained by sewing together one-soliton BPS fields at finite separation. Section 2.1 explicitly says that two-soliton initial data are constructed by sewing one-soliton fields, and such sewn configurations are not static solutions of the full equations: they contain radiation from the seam and from the finite-distance interaction. The slow motions observed in Sections 2–4 therefore cannot be used as evidence for the zero-mode mechanism without an estimate of the size of the sewn-data error. Section 5.2 states that 'BPS fields' were determined numerically, but no procedure is given for how the numerical BPS solutions were obtained, how their errors were estimated, or how the initial data were prepared for the long-time runs; this matters because the central claim is that tiny initial errors, not physical forces, drive the drift.","section":"§2.1 and §5.2"}],"minor_comments":[{"comment":"The BPS bound for the energy in Eq. (2.8) is E(λ) = 16√((2−λ)/(2+λ)), not 8√((2−λ)/(2+λ)); at λ=0 this gives 16, matching the two locked kinks, whereas the stated value 8 is the single-field result. The comparison with 'E(λ=1.0)≈4.618' is therefore off by a factor of 2 and should be rechecked.","section":"§2.1.1, Eq. (2.8)"},{"comment":"There are several typos in the abstract and in Section 5.3: 'exitations', 'unevoidable', 'more that one zero mode', and 'inavoidable' should be corrected; similar typos appear throughout the text (e.g., 'fololows', 'looses', 'controlls', 'useed').","section":"Abstract and §5.3"},{"comment":"The energy conservation claims are stated inconsistently: the abstract says '10^-5%' while Section 4.1 reports 'up to 10^-7' for one simulation, and some energy plots appear to show box-energy rather than total energy. The diagnostic used for each stated conservation level should be defined precisely.","section":"§4.1 and §5.2"},{"comment":"The statement that for λ=0 the decoupled system 'sends out only radiation' is supported by a citation to a discrete sine-Gordon paper, but the relevance of that result to the present continuum setup is not explained. Since the λ=0 case is used as the baseline for the zero-mode argument, a brief derivation or explicit numerical demonstration would be helpful.","section":"§5.3"},{"comment":"Reference [6] has an incomplete arXiv identifier ('arXiv: 1206.447' is missing digits), and reference [8] is missing a period. In Fig. 24 the y-axis label '4.' should read '4.0', and in the caption of Fig. 28 'λ−1.95' should read 'λ=−1.95'.","section":"References and figures"}],"recommendation":"major_revision","confidential_remarks":"This is a purely numerical paper, but it contains no numerical methods section, no code or data availability statement, and no convergence study. The central claim depends on distinguishing numerical-error-driven zero-mode drift from discretization artifacts, and the manuscript's own admission in Section 5.4 that the sign dependence is unexplained strengthens the need for additional evidence. The paper is within the journal's scope, but it needs a substantial revision that adds reproducibility details and direct diagnostics for the zero-mode mechanism before it can be recommended for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the takeaway: this paper reports a genuinely new numerical observation. Multi-field BPS static solutions, when evolved with the full equations, slowly drift along internal zero modes because tiny numerical errors excite those modes. That is not something the authors' earlier paper on BPS systems discussed, and it is a real caution for anyone doing long-time simulations of BPS solitons. The effect is small—positions change by hundredths to a half over thousands of time units—but it is systematic, with the sign of the drift depending on the sign of λ and on whether you are looking at kink-kink or kink-antikink sectors.\n\nThe paper does the honest empirical work: many long runs, energy conserved to 10^-7 relative, careful tracking of soliton positions, and the λ=0 control case behaves as expected. The authors also plainly admit what they cannot explain, most notably the sign dependence of the drift. That is good scientific practice, and the central caution does not depend on the explanation being complete.\n\nThe soft spots are real but in proportion. The zero-mode explanation is qualitative. The paper never measures the projection of the initial numerical error onto the two zero modes, nor does it do a convergence study, a lattice-spacing scan, or a single- vs double-precision comparison. The global energy check is blind to motion along the flat direction, so it does not by itself rule out a deterministic discretization artifact. The sewn initial conditions could in principle inject a small systematic kick rather than random roundoff. The authors do not supply code or data. These are exactly the things a referee should ask for. None of them, though, undermines the raw observation that these 'static' solutions do not stay put; they only leave the mechanism open.\n\nThe scattering phenomenology—kink repulsion, kink-antikink attraction, breather/oscillon formation—is mostly an expected extension of sine-Gordon physics, and the bound-state interpretation in Section 2.1.1 is a nice but minor observation. The paper's real payload is the zero-mode drift warning.\n\nWho should read it: anyone doing numerical work with BPS systems in more than one scalar field. It deserves a serious referee, not a desk rejection. I would recommend conditional acceptance: ask the authors for a convergence study, a precision comparison, and ideally code/data, or at least a direct projection of the initial error onto the zero modes. Without that, the mechanism remains a plausible hypothesis rather than a demonstrated one.","headline":"Careful numerics suggest multi-field BPS solitons drift along zero modes under tiny numerical errors; the claim is plausible and worth taking seriously, though the mechanism is not yet pinned down.","tokens_in":20348,"tokens_out":2439,"would_cite":true,"duration_ms":25537,"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":"The paper claims that in two-field BPS systems, tiny unavoidable numerical errors excite a relative zero mode, so ostensibly static soliton solutions drift apart or together very slowly, with direction and rate controlled by the coupling…","keywords":["BPS systems","zero modes","Sine-Gordon model","solitons","kinks","antikinks","coupled scalar fields","numerical simulation"],"falsifier":"Run the same static BPS initial data at successively smaller lattice spacings (or with successively smaller time steps) and measure the drift speed of the soliton positions; if the extrapolated speed does not tend to zero with the numerical error, the drift is a discretization artefact rather than zero-mode excitation. Alternatively, compute the projection of the initial numerical error onto the relative zero mode and check that the observed drift velocity agrees with it.","tokens_in":19434,"feed_emoji":"🌊","tokens_out":9591,"duration_ms":87536,"temperature":0.7,"pith_summary":"This paper studies a (1+1)-dimensional model of two coupled Sine-Gordon-like scalar fields whose one-soliton solutions satisfy BPS (Bogomol'nyi–Prasad–Sommerfield) equations, with an interaction strength $\\lambda$ satisfying $|\\lambda|<2$. The central claim is that ostensibly static BPS solutions of such multi-field systems are not static in long numerical evolutions: tiny numerical errors excite a second zero mode, causing the solitons to drift apart or together extremely slowly, even though total momentum conservation cancels the overall translation. The direction of the drift flips with the sign of $\\lambda$ and with whether the system is kink-kink or kink-antikink. Understanding this zero-mode artefact matters because it can contaminate studies of multi-soliton interactions, whose exponential forces are often just as slow as the spurious drift. The same mechanism is offered as an explanation of the slowly oscillating kink-antikink configurations seen in the two-breather systems studied earlier in the paper.","feed_headline":"Tiny numerical errors move 'static' BPS solitons","feed_subtitle":"A hidden zero mode lets small numerical noise cause slow attraction or repulsion in coupled two-field models.","key_machinery":"The load-bearing object is the BPS sector defined by the prepotential $U=-4(\\cos\\phi_1+\\epsilon\\cos\\phi_2)$ together with the modified gradient flow $\\nabla_{\\gamma}U=\\frac{4}{4-\\lambda^2}(2\\partial_{\\phi_1}U+\\lambda\\partial_{\\phi_2}U,\\lambda\\partial_{\\phi_1}U+2\\partial_{\\phi_2}U)$. The first-order BPS equations are $\\partial_x\\phi_1=\\frac{4}{4-\\lambda^2}(4\\sin\\phi_1+2\\lambda\\epsilon\\sin\\phi_2)$ and $\\partial_x\\phi_2=\\frac{4}{4-\\lambda^2}(2\\lambda\\sin\\phi_1+4\\epsilon\\sin\\phi_2)$. Because two first-order equations in one dimension leave two constants free, the one-soliton solution carries two zero modes: a global translation and a relative shift of the two fields' soliton positions. It is the relative mode that numerical noise excites, and this is the mechanism that converts an ostensibly static solution into a slow drift. The single-field reduction $\\phi_2=\\phi_1+\\pi$ yields the energy formula $E(\\lambda)=8\\sqrt{(2-\\lambda)/(2+\\lambda)}$ used to explain radiation from bound states.","core_discovery":"The paper's central discovery, stated in its own terms, is that a BPS solution of a system with two scalar fields and $|\\lambda|<2$ has two zero modes, not one: because the BPS equations are first order, the solution depends on two constants, which show up as an overall translation and as a relative motion of the solitons of the two fields. When the equations are evolved numerically, unavoidable round-off error excites the relative zero mode, while momentum conservation kills the translational one. As a result, the 'static' BPS solution evolves very slowly — the paper reports position changes of order $0.05$ to $0.5$ over runs of $10^3$ to $10^4$ time units — with repulsion for kink-antikink systems at $\\lambda>0$, attraction at $\\lambda<0$, and the opposite signs for two-kink systems, at a rate that grows with $|\\lambda|$. The paper also shows that a locked bound state of the two fields can be reinterpreted as a single Sine-Gordon soliton with energy $E(\\lambda)=8\\sqrt{(2-\\lambda)/(2+\\lambda)}$, which explains why such bound states radiate their excess energy.","pith_inferences":["A convergence study across lattice spacings or time steps would test the mechanism: if the drift is numerical noise exciting a zero mode, its speed should shrink with the numerical error, something the paper does not report.","One could compute the projection of the numerical error of the initial fields onto the relative zero mode and compare it with the measured drift velocity; if they do not match, the drift has a different source.","The sign flip between kink-kink and kink-antikink systems might follow from a collective-coordinate effective potential built on the two zero modes; the paper leaves this derivation open.","If the mechanism is generic, the same slow drift should be observable in other multi-field BPS models, such as coupled $\\phi^4$ or baby-Skyrme-type systems, providing a cheap numerical test of the claim."],"forward_implications":["Long-time numerical evolutions of multi-field BPS systems cannot be assumed to remain static; the observed slow motions can be numerical artefacts caused by excitation of the relative zero mode.","For kink-antikink systems, $\\lambda>0$ gives a slow repulsion and $\\lambda<0$ a slow attraction; for two-kink systems the signs are reversed, and the drift speed increases with $|\\lambda|$.","Studies of multi-soliton interactions at large separation, where the real forces are exponentially small, will be contaminated by this zero-mode drift.","A locked two-field bound state can be described as one Sine-Gordon soliton with excess energy $E(\\lambda)=8\\sqrt{(2-\\lambda)/(2+\\lambda)}$, which explains the late-time radiation and breather emission seen in the simulations.","The authors expect the same zero-mode artefact in other multi-field BPS solitonic systems, making the warning generic rather than specific to this model."],"supporting_citations":[{"why":"Introduces the two-field model, its Lagrangian and BPS equations, and the one-soliton solutions whose static character is tested here.","marker":"[1]"},{"why":"Supplies the decoupled $\\lambda=0$ result that numerical noise in a discrete sine-Gordon kink produces radiation but no soliton motion, motivating the role of momentum conservation and extra zero modes.","marker":"[10]"},{"why":"Underlies the concluding claim that the models are close to integrable, which supports interpreting the long-time behaviour as quasi-integrable.","marker":"[5]"}],"fun_headline_variants":["Zero modes turn 'static' BPS solitons into slow movers","Numerical noise excites hidden mode, drifts BPS solitons","BPS solitons stir: round-off wakes relative zero mode","Tiny errors make 'static' BPS solitons drift slowly","Coupled kinks move: zero mode feeds on round-off error"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole explanation rests on the assumption that the slow motions are caused by numerical noise exciting the relative zero mode, rather than by the discretization scheme or by radiation generated when the initial data are sewn together; the paper reports only global energy conservation to $10^{-5}\\%$, not a direct measurement of the numerical error in the modes.","fun_headline_variants_meta":{"raw":{"variants":["Zero modes turn 'static' BPS solitons into slow movers","Numerical noise excites hidden mode, drifts BPS solitons","BPS solitons stir: round-off wakes relative zero mode","Tiny errors make 'static' BPS solitons drift slowly","Coupled kinks move: zero mode feeds on round-off error"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000626,"raw_usage":{"total_tokens":2960,"prompt_tokens":1072,"completion_tokens":1888,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":688,"completion_tokens_details":{"reasoning_tokens":1792}},"tokens_in":688,"tokens_out":1888,"duration_ms":48492,"temperature":1.0,"reasoning_tokens":1792,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:53:40.326808+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same static BPS initial data at successively smaller lattice spacings (or with successively smaller time steps) and measure the drift speed of the soliton positions; if the extrapolated speed does not tend to zero with the numerical error, the drift is a discretization artefact rather than zero-mode excitation. Alternatively, compute the projection of the initial numerical error onto the relative zero mode and check that the observed drift velocity agrees with it.","supporting_citations":[{"cited_title":"Some Comments on BPS systems","cited_arxiv_id":"1803.08985","evidence_quote":"Introduces the two-field model, its Lagrangian and BPS equations, and the one-soliton solutions whose static character is tested here."},{"cited_title":"Spontaneous emission of radiation from a discrete sine-Gordon kink","cited_arxiv_id":null,"evidence_quote":"Supplies the decoupled $\\lambda=0$ result that numerical noise in a discrete sine-Gordon kink produces radiation but no soliton motion, motivating the role of momentum conservation and extra zero modes."},{"cited_title":"The concept of quasi-integrability: a concrete example","cited_arxiv_id":"1011.2176","evidence_quote":"Underlies the concluding claim that the models are close to integrable, which supports interpreting the long-time behaviour as quasi-integrable."}],"review_version":1}