{"id":"f714e641-0841-4182-bc39-03753f8e30f5","arxiv_id":"1908.07499","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Micromorphic continuum elasticity for critical mechanical metamaterials yields a topological invariant equal to the difference in zero-energy edge modes on opposing boundaries, with decay lengths set by the square of the surface wavelength.","lead":"Flexible periodic materials made of springs or hinges can have very soft, low-energy modes that concentrate on surfaces. This paper builds a continuous elastic description of such materials and derives a new topological rule that predicts how many of these soft modes sit on one edge versus the opposite edge, including how deep into the material they reach.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (17) counts only truncated-det zeros; excluding short-wavelength modes by decree, not by lattice argument, leaves NL−NR unproven.","rationale":"The paper's central claim is a long-wavelength bulk-boundary correspondence: Eq. (17) computes an integer from the continuum determinant and asserts it equals the difference in zero-energy edge modes on two opposing edges. For this to be true, the zeros of det R enclosed by the contour must be exactly the physical long-wavelength edge modes, with all other zeros irrelevant. The derivation does not establish this. It identifies two modes of the form qx=α± qy + iβ± qy² and labels the remaining roots of the third-order truncation as 'short-wavelength modes, non-physical and dictated by the order at which we terminate our expansion.' This is a decision about which zeros are physical, not a derivation from the lattice theory. Appendix D does not close the gap: it analyzes the contribution of a single zero z−z0, shows the canceled curved-portions contribute O(ε²/r), and chooses r=ε^{1/2} to suppress spurious-mode errors, but it assumes the contour already encloses all long-wavelength zeros and does not prove that the exact lattice determinant has no additional enclosed zeros. The paper's own statement about noticeable numerical error near the topological transition is consistent with this being the fragile step. The reader's weakest_assumption identifies the same load-bearing step: the truncation and exclusion of short-wavelength modes. Because this is a missing justification rather than a demonstrated internal contradiction, the existing CONDITIONAL verdict is the right one; I am not moving it. I also note that the paper provides no machine-checked proof and no released code, so the numerical demonstration cannot independently settle the question. A direct comparison of Eq. (17) with exact lattice winding and finite-strip edge-mode counts would settle whether the concern lands.","tokens_in":15182,"tokens_out":20816,"duration_ms":242448,"concrete_test":"For the generalized-kagome family g(x) used in Fig. 4, build the exact Bloch rigidity matrix of the full spring network (no continuum truncation) and compute: (1) the standard full-Brillouin-zone winding of det R at small qy values; (2) the partial integral in Eq. (17) using the exact determinant; and (3) the actual zero-energy edge-mode imbalance NL−NR from finite-strip diagonalization with periodic boundary conditions in y, for several widths W and qy=ε with W ≫ 1/(β ε²). If (1), (2), and (3) agree as ε→0, the truncation concern is resolved; if (2) or (3) differs from the lattice edge-mode count, discarded short-wavelength zeros are load-bearing and the central correspondence fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central equality NL−NR = Eq. (17) rests on the claim that every physical edge mode relevant to the imbalance is captured by low-q complex zeros of det R of the form qx=α± qy + iβ± qy², and that all O(qy⁰) roots of the truncated determinant are non-physical artifacts (Sec. V, immediately after Eqs. (15)–(16)). This exclusion is asserted, not proved: the text says the continuum formulation 'deliberately excludes' short-wavelength modes, but the claim concerns the physical lattice, whose edge-mode count can in principle include finite-qx zero modes. Appendix D controls the contour error only for a single model factor z−z0 and explicitly assumes that the contour encloses all long-wavelength zeros; it does not show that the full lattice rigidity map has no additional zeros inside the ε→0 contour, nor that such zeros would cancel in NL−NR. If the exact determinant has extra enclosed zeros at intermediate scales (e.g. qx∼ε^{2/3}) or if the q-dependent self-stress basis introduces additional phase winding, Eq. (17) can differ by an integer from the actual edge-mode imbalance. The paper's own report of noticeable numerical error near the topological transition marks the place where this genericity assumption is least secure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a micromorphic continuum elasticity for flexible mechanical metamaterials, starting from a bond-extension expansion in strain and strain gradients. It derives bulk and surface energy terms, constructs a square rigidity map R after projecting out intra-cell relaxations via states of self-stress, and identifies zero-energy edge modes with complex-wavevector zeros of det R of the form qx = α± qy + i β± qy². The central claim is Eq. (17), which equates the edge-mode imbalance NL−NR on two opposing edges with the winding of det R along a long-wavelength contour, and Eq. (24), which predicts that the decay length of these boundary modes scales as the square of the surface wavelength. The paper also introduces soft directions, studies how the polarization changes with interface orientation, and discusses experimental length and energy scales.","tokens_in":15432,"tokens_out":4730,"duration_ms":48365,"significance":"If the central equality holds, the paper provides a substantive bridge between discrete topological mechanics of Maxwell lattices and continuum micromorphic elasticity, with concrete, falsifiable predictions about boundary-mode decay lengths and an experimentally accessible polarization invariant. Strengths include the explicit construction of the rigidity map from the microstructure, the careful separation of bulk and surface energies, the use of the argument principle with a controlled O(ε^(1/2)) contour error in Appendix D, and the numerical demonstration on a generalized kagome family. The paper also usefully identifies soft directions and the geometric suppression of decay lengths. However, the main bulk-boundary correspondence is proven only for a truncated determinant under an explicit genericity assumption about the location of its zeros, and the paper states this assumption rather than deriving it from the lattice theory. The result is therefore conditional, though plausibly repairable.","major_comments":[{"comment":"The equality NL−NR = (17) rests on the assertion that every physical long-wavelength edge mode corresponds to a low-q complex zero of det R of the form (16) and that all other zeros of the truncated determinant are non-physical artifacts. This exclusion is introduced by decree ('our continuum formulation deliberately excludes them') rather than derived from the lattice rigidity matrix. Appendix D bounds the contour error only for a single factor z−z0, and its error estimate is stated under the assumption that the contour 'actually encloses all of the long-wavelength zero modes'; it does not show that the exact determinant has no additional zeros inside the ε→0 contour (e.g., at intermediate scales qx∼ε^(2/3)) or that such zeros would cancel in NL−NR. Since the right-hand side of Eq. (17) counts only enclosed zeros of the truncated determinant, any such extra or excluded zero changes the result by an integer. The authors should supply either a lattice-level argument showing that the edge-mode imbalance equals the winding of det R around a contour excluding all other zeros, or a direct numerical check against the lattice topological invariant [15] for the generalized kagome family.","section":"§V, immediately after Eq. (16), and Appendix D"},{"comment":"The determinant expansion is truncated at third order in q, and the paper states that terminating the expansion to order n 'indicates the presence of n zero modes,' with two taking the long-wavelength form (16) and the rest being short-wavelength artifacts. For fixed qy=ε, the truncated determinant is a cubic in qx, so the root count depends on the truncation order; the claim that the two physical roots are precisely those of the form (16) is not justified by an error estimate uniform in qy. The observation in Fig. 4(a) of 'noticeable error' very close to the topological transition is exactly the regime in which the separation between long-wavelength and short-wavelength roots is no longer controlled, and the manuscript does not explain how an integer-valued imbalance is recovered there.","section":"§V, Eqs. (15)–(17)"},{"comment":"The error estimate O(ε^(1/2)) is derived for the phase change of a single zero z−z0, under the hypotheses r≫|z0| and that the contour encloses all long-wavelength zeros. The appendix itself notes that if this enclosure condition is not met, the error 'increases abruptly to O(1).' This is a limitation of the proof as written, not a mere technical remark: the main result (17) is therefore not yet a theorem about the physical lattice, but a genericity assumption about the zero set of det R. Stating this as a theorem with explicit hypotheses and verifying those hypotheses numerically for the lattices considered would substantially strengthen the paper.","section":"Appendix D, Eq. (41)"}],"minor_comments":[{"comment":"The normalization in the definition of q̂± appears to be missing a square: the denominator should be sqrt(1+α±²), not sqrt(1+α±).","section":"Eq. (20)"},{"comment":"The notation A''_{3,0}(0,3) is unexplained and appears to be a typo; please clarify whether this denotes a coefficient of the rotated determinant or a typographical artifact.","section":"Eq. (22) and surrounding text"},{"comment":"The phrase 'surface surface terms' appears twice; one occurrence should be corrected.","section":"Secs. I and III"},{"comment":"The text states that the parameters entering the decay-length expression 'can't be measured by the bulk response,' yet the same section claims these are macroscopic experimental observables. Please clarify which quantities in Eq. (24) are independently measurable and which require knowledge of the microscopic rigidity map.","section":"Sec. VI, Eq. (24)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a promising approach, but the central bulk-boundary correspondence is conditional on an unverified genericity assumption about the zeros of the determinant of the rigidity map. The missing lattice-level argument or numerical verification appears feasible within the paper's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe paper you'll want to know about: Saremi and Rocklin give the first genuine continuum treatment of topological polarization in Maxwell-coordinated mechanical lattices. The new pieces are real: the micromorphic rigidity map built from self-stress projection, the strain-gradient surface energy terms, the winding invariant of Eq. (17) relating bulk structure to the difference in edge-mode counts, and the lambda^2 decay length prediction. None of those appear in the lattice Kane–Lubensky line of work they cite. The presentation is clear and the authors are upfront that ref. [32] appeared independently.\n\nWhere it gets soft: the central equality NL−NR = Eq. (17) depends on the claim that all physical edge modes relevant to the imbalance are captured by low-q complex zeros of det R(q) of the form qx = α± qy + iβ± qy^2, and that all O(qy^0) roots of the truncated determinant are non-physical artifacts. The paper says the continuum formulation 'deliberately excludes' short-wavelength modes, but the quantity NL−NR is a statement about the physical lattice. Appendix D controls the contour error only for a single model factor z−z0 and explicitly assumes the contour encloses all long-wavelength zeros; it doesn't show the full lattice rigidity map has no extra enclosed zeros, nor that such zeros cancel. So the bulk-boundary correspondence is conditional on a genericity assumption that is plausible but not proved. The authors do flag the numerical error near the transition, which is telling.\n\nThat said, the paper does not fit constants or tune anything to get the invariant; the coefficients come from the microstructure. The circularity burden is low. The numerical evidence is qualitative rather than a reproducible quantitative check—no code, no error bars—but for a theory paper that's a weakness in presentation, not a fatal one.\n\nMy take: this is a solid, imaginative piece worth serious refereeing. The main theorem needs either a stronger lattice-level argument or a clear statement that the continuum invariant predicts the imbalance only when the truncation is valid. I'd send it to review with a request for that clarification, not desk-reject it.\n\nBest","headline":"A genuine continuum analogue of lattice topological polarization with a new invariant and decay scaling, but the main equality is conditional on a truncation that excludes short-wavelength modes by decree rather than proof.","tokens_in":15910,"tokens_out":1893,"would_cite":true,"duration_ms":19214,"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":"This paper claims that in flexible metamaterials, a bulk winding number gives the difference in zero-energy modes on two opposite edges and sets the edge-mode depth by the square of surface wavelength.","keywords":["topological mechanics","flexible mechanical metamaterials","micromorphic elasticity","isostatic lattices","zero-energy edge modes","topological polarization","strain-gradient elasticity","kagome lattice"],"falsifier":"Compute the full determinant of $R(q)$ for the generalized kagome family without truncating at third order: if the winding evaluated at $q_y=\\epsilon$ over $|q_x|\\leq \\sqrt{\\epsilon}/|l_1|$ differs from the true $N_L-N_R$ obtained by diagonalizing a finite sample, the truncation discards physical modes. Equivalently, build a finite kagome sample with polarization $\\Delta N=2$, impose a boundary distortion of wavelength $\\lambda$, and check whether the mode penetrates a depth of order $(\\lambda/|l_1|)^2|l_1|$ rather than order $\\lambda$; a Rayleigh-like depth would falsify Eq. (24).","tokens_in":14959,"feed_emoji":"⚙️","tokens_out":11884,"duration_ms":106423,"temperature":0.7,"pith_summary":"This paper argues that the topological edge-mode physics previously found in discrete critically coordinated lattices survives when the lattice is coarse-grained into a micromorphic continuum. It constructs a bulk winding number, Eq. (17), that equals the difference $N_L - N_R$ in long-wavelength zero-energy modes localized on two opposing edges, and shows that these modes exist on a mesoscopic length scale: their decay length grows as the square of the surface wavelength, Eq. (24). The claim matters because it turns a lattice-level topological property into a macroscopic observable of flexible mechanical metamaterials, accessible through strain-gradient elasticity rather than atomic-scale enumeration.","feed_headline":"Flexible metamaterials' edge modes obey a continuum topological rule","feed_subtitle":"A strain-gradient invariant fixes which edge yields; mode depth grows as the square of surface wavelength.","key_machinery":"The load-bearing object is the relaxed rigidity map $R_{m,ij}(q)$, a square matrix obtained from the initial bond-extension map after projecting out short-wavelength relaxations onto the states of self stress, so that it maps the $d$ independent smooth strain components to $d$ constraints (the continuum condition for critical coordination). Its determinant $\\det R(q)$ vanishes exactly at zero-energy modes; allowing $q$ to be complex turns edge localization into zeros at $q_x = \\alpha_\\pm q_y + i\\beta_\\pm q_y^2$. The machinery is the contour integral (Eq. (17)) over real $q_x$ at fixed small imaginary $q_y=\\epsilon$, which by the argument principle counts left-edge minus right-edge long-wavelength modes while the curved parts of the contour cancel; the strain-gradient surface energy term, Eq. (5), provides the physical boundary energetics that makes these modes cost zero energy on one edge.","core_discovery":"The central claim is that in the micromorphic limit of a critically coordinated flexible structure (equal numbers of smooth strain constraints and strain degrees of freedom), the difference between the numbers of long-wavelength zero-energy modes on two opposing edges is a bulk topological invariant. The invariant is the winding of $\\arg \\det R(q_x, q_y=\\epsilon)$ as $q_x$ runs over a real interval whose width shrinks as $\\sqrt{\\epsilon}$: $$N_L - N_R = \\frac{1}{\\pi}\\lim_{\\epsilon\\to 0^+}\\int_{-\\sqrt{\\epsilon}/|l_1|}^{\\sqrt{\\epsilon}/|l_1|} dq_x\\, \\partial_{q_x}\\arg\\det R(q_x, q_y=\\epsilon),$$ with zeros of $\\det R$ in the upper half-plane counted as left-edge modes and those in the lower half-plane as right-edge modes. The authors show that this continuum invariant is quantized exactly in the long-wavelength limit, that it varies with interface orientation and jumps when the normal crosses a soft direction, and that the corresponding edge modes decay into the bulk over a length $\\zeta_\\pm \\sim (\\lambda/|l_1|)^2 |l_1|$ set by the surface wavelength $\\lambda$. In their picture the same strain-gradient surface terms that produce the edge energies are what break inversion symmetry and support the polarization.","pith_inferences":["A direct test would be to measure the effective edge stiffness of a 3D-printed critically coordinated lattice as a function of imposed surface wavelength: if the paper is right, the boundary mode depth should scale as $\\lambda^2$, whereas a conventional Rayleigh-type surface mode scales as $\\lambda$, making the two contributions separable.","The same winding integral could be evaluated from measured strain-gradient elastic coefficients rather than from a lattice model, which would let experiments determine $N_L-N_R$ before any edge is loaded.","If the contour cancellation is robust, the polarization should survive surface disorder and rounding: modifying the boundary termination changes the short-wavelength zeros but not the difference $N_L-N_R$, a stability that finite lattices would show in numerics."],"forward_implications":["The bulk microstructure generates boundary elastic terms in the continuum energy: the strain-gradient part of the energy reduces to a surface integral, so edges soften or stiffen independently of the bulk response.","The continuum topological invariant is quantized in the long-wavelength limit and reproduces the lattice polarization $\\Delta N = N_L - N_R$ without requiring a Brillouin zone.","Topological edge modes are mesoscopic: for a boundary distortion of wavelength $\\lambda$ the mode penetrates a depth of order $(\\lambda/|l_1|)^2 |l_1|$ cells, so the effect survives at scales far above the unit cell but below system size.","As the interface normal is rotated, $\\Delta N(\\theta_n)$ changes only when the normal crosses a soft direction, giving a directional polarization that can be mapped in experiments.","Because the surface theory uses only bulk strain-gradient coefficients and bond geometry, the predicted edge-mode imbalance can be observed even when the microscale structure is not resolved."],"supporting_citations":[{"why":"It supplies the micromorphic continuum elasticity framework in which strain gradients and internal degrees of freedom are retained in the energy.","marker":"[14]"},{"why":"It introduces the discrete-lattice topological invariant for critically coordinated lattices that the present continuum invariant generalizes.","marker":"[15]"},{"why":"It provides the projection onto states of self stress used to define the relaxed rigidity map after short-wavelength relaxation.","marker":"[16]"},{"why":"It supplies the rigidity and equilibrium matrix formalism and the determinant-based topological analysis that Eq. (17) adapts to the continuum.","marker":"[22]"},{"why":"It gives the index theorem that fixes the number of self-stress states in the critical-coordination argument.","marker":"[23]"},{"why":"It provides the zero-energy strain modes whose low-wavevector zeros make $\\det R$ start at $O(q^2)$ and shape the long-wavelength edge-mode dispersion.","marker":"[26]"},{"why":"It gives the soft-direction characterization of the kagome lattice that determines where the polarization $\\Delta N(\\theta_n)$ switches as the interface rotates.","marker":"[29]"}],"fun_headline_variants":["Continuum topology dictates edge modes in flexible metamaterials","New invariant sets edge-mode imbalance in soft lattices","Edge-mode depth scales as wavelength squared in flexible structures","Bulk topology pins which edge yields in flexible metamaterials","Winding number fixes edge-mode difference in micromorphic elasticity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"At bottom, the argument requires that the long-wavelength expansion of the determinant, cut off at third order in wavevector, already captures every mode that can live on an edge; if any short-wavelength lattice mode contributes to the winding count, Eq. (17) stops counting the true edge-mode imbalance.","fun_headline_variants_meta":{"raw":{"variants":["Continuum topology dictates edge modes in flexible metamaterials","New invariant sets edge-mode imbalance in soft lattices","Edge-mode depth scales as wavelength squared in flexible structures","Bulk topology pins which edge yields in flexible metamaterials","Winding number fixes edge-mode difference in micromorphic elasticity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000545,"raw_usage":{"total_tokens":2615,"prompt_tokens":959,"completion_tokens":1656,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":575,"completion_tokens_details":{"reasoning_tokens":1577}},"tokens_in":575,"tokens_out":1656,"duration_ms":12244,"temperature":1.0,"reasoning_tokens":1577,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:18:42.366095+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full determinant of $R(q)$ for the generalized kagome family without truncating at third order: if the winding evaluated at $q_y=\\epsilon$ over $|q_x|\\leq \\sqrt{\\epsilon}/|l_1|$ differs from the true $N_L-N_R$ obtained by diagonalizing a finite sample, the truncation discards physical modes. Equivalently, build a finite kagome sample with polarization $\\Delta N=2$, impose a boundary distortion of wavelength $\\lambda$, and check whether the mode penetrates a depth of order $(\\lambda/|l_1|)^2|l_1|$ rather than order $\\lambda$; a Rayleigh-like depth would falsify Eq. (24).","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the micromorphic continuum elasticity framework in which strain gradients and internal degrees of freedom are retained in the energy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It introduces the discrete-lattice topological invariant for critically coordinated lattices that the present continuum invariant generalizes."},{"cited_title":"Zhang and X","cited_arxiv_id":null,"evidence_quote":"It supplies the rigidity and equilibrium matrix formalism and the determinant-based topological analysis that Eq. (17) adapts to the continuum."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the zero-energy strain modes whose low-wavevector zeros make $\\det R$ start at $O(q^2)$ and shape the long-wavelength edge-mode dispersion."}],"review_version":1}