{"id":"073c5a80-342a-432e-b8a9-9eb193a66933","arxiv_id":"2602.09335","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In the half-filled Haldane-Holstein model, electron-phonon coupling drives a first-order transition from a Chern insulator to a staggered charge-density wave at gc/t1≈1.6 (t2/t1=0.2, ω0/t1=1).","lead":"Using quantum Monte Carlo, this paper shows that in a model combining a Chern insulator with vibrating atoms (phonons), strong coupling drives an abrupt first-order transition into a charge-ordered state. It maps out the phase boundary and identifies measurable signatures such as a collapse of topological markers and boundary states.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sign-problem control at the transition is asserted but not quantified; without ⟨sign⟩ and error bars at L=15, the first-order CI-CDW transition and marker collapse could be statistical artifacts.","rationale":"The reader's weakest assumption correctly identifies the unresolved sign-problem control as the most load-bearing point. My independent reading of the text confirms that the central first-order transition claim is supported by DQMC simulations near a sign minimum, with no quantitative sign values, no effective sample sizes, and no error bars on the key observables (Sc/N, Bott index, Chern marker, energy). The SM's admission that DQMC is unfeasible at ω0/t1>4 strengthens the need for such a validation. I considered whether a different concern—e.g., the approximate nature of Green's-function-based topological markers—might be more fundamental, but the authors themselves note the markers degrade near the transition, and the phase-coexistence/energy-drop diagnostics are independent of the markers. The sign problem is therefore the single condition that, if unsatisfied, would undermine all diagnostics. Since this is the same assumption the reader flagged, and the requested verification is already implicit in the conditional verdict, I recommend no change to the reader's verdict: CONDITIONAL remains appropriate pending quantitative sign/error-bar validation.","tokens_in":17209,"tokens_out":3101,"duration_ms":31193,"concrete_test":"Reanalyze the raw DQMC data at t2/t1=0.2, ω0/t1=1, g/t1=1.6 for L=15, T/t1=1/30: report ⟨sign⟩ with its statistical error and the effective sample size per histogram bin (N_eff = N_meas × ⟨sign⟩²) for the Sc reweighted histogram. Also compute Sc/N with jackknife error bars at g/t1=1.55, 1.60, 1.65 for L=6,9,12,15. If the L=15 error bars at g/t1=1.6 overlap both the g=1.55 and g=1.65 values, or if N_eff<100 in any bin of Fig. 2(e), the bimodality and abrupt jump cannot be distinguished from sign-noise artifacts. If instead ⟨sign⟩≥0.2 with N_eff>1000 and the error bars remain small, the first-order interpretation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—an abrupt first-order CI-CDW transition at g/t1≈1.6 for t2/t1=0.2, ω0/t1=1—rests on DQMC data taken at the point where the complex-weight sign problem is worst. Figure 2(b) shows a pronounced ⟨sign⟩ dip near g/t1≈1.6, and the largest system (L=15, T/t1=1/30) is admitted to have the smallest sign, yet no numerical values, error bars, or effective sample sizes are reported anywhere in the main text or SM. Without these, the reweighted histograms in Fig. 2(d–f) are not self-validating: if ⟨sign⟩ within a bin is small (e.g., <0.1), the reweighted P(Sc) has enormous variance, and a bimodal shape can be produced by sign fluctuations that correlate with Sc, not by true phase coexistence. The authors cite analogous sign behavior at CDW transitions [60,61], but that does not quantify their own error budget. The Bott index and local Chern marker are also constructed from the equal-time Green's function, which is directly degraded by sign noise near the transition; the observed collapse of topological markers could be premature if G is statistically unreliable. The SM openly states DQMC becomes unfeasible at ω0/t1>4 due to the sign problem, making it especially important to demonstrate quantitatively that the low-frequency regime used here is actually controlled. Without reporting ⟨sign⟩, N_eff, and error bars on Sc/N, B, and C at the reported parameters, the abrupt jump and bimodality cannot be distinguished from uncontrolled reweighting noise.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a determinant quantum Monte Carlo (DQMC) study of the half-filled Haldane-Holstein model on a honeycomb lattice, with complex next-nearest-neighbor hopping t2=0.2t1 and phonon frequency ω0=t1. The central claim is that increasing electron-phonon coupling g drives an abrupt, first-order transition from a Chern insulator (CI) to a staggered charge-density wave (CDW), with the CDW acting as a dynamical sublattice (Semenoff) mass. The transition is identified at g/t1≈1.6 via the simultaneous collapse of the many-body Bott index and local Chern marker, a sharp rise in the charge structure factor, bimodal reweighted histograms of Sc, a drop in the total energy, and a minimum in the average fermion sign. Spectral functions and open-boundary LDOS are used to support gap closing/reopening and loss of edge states. The paper also maps the phase diagram in (t2/t1, g/t1) and in (λ, t1/ω0), with mean-field and exact-diagonalization checks in the adiabatic and antiadiabatic limits. The authors deposit their data and make no free-parameter fits.","tokens_in":17593,"tokens_out":4221,"duration_ms":43282,"significance":"If the central claim is correct, this is a valuable unbiased numerical demonstration that retarded electron-phonon coupling can destroy Chern topology via a first-order transition into a staggered CDW, with direct relevance to TMDs and layered quantum Hall systems. The combination of independent diagnostics — charge correlations, topological markers, spectral functions, and the fermion sign — is conceptually appealing, and the lack of fitted parameters is a strength. The paper also offers a falsifiable prediction: the average determinant sign dips precisely at the phase boundary, which is an interesting empirical observation consistent with earlier work. However, the statistical control of the DQMC simulation, especially at the transition point where the sign problem is worst, is not demonstrated quantitatively. Since the central first-order claim rests on histograms and energy drops at finite size, the significance will be fully established only after the missing error analysis and sign-problem validation are provided.","major_comments":[{"comment":"The central claim is not yet statistically supported because the sign problem at the transition is asserted to be mild but is never quantified. The paper reports no values of ⟨sign⟩, no error bars, no effective sample sizes, and no autocorrelation times. The data shown in Fig. 2(b) and the reweighted histograms in Fig. 2(d)–(f) are obtained precisely in the coupling window where the sign is minimal. If within a reweighted bin ⟨sign⟩ is small, the variance of P(Sc) becomes large and a bimodal shape can be produced by sign fluctuations correlated with Sc rather than by genuine phase coexistence. The SM admits that for ω0/t1=4 the sign problem forces βt1 to be halved, and that ω0/t1>4 is unfeasible; this makes the low-frequency claim load-bearing. I request a quantitative reporting of ⟨sign⟩, N_eff, and error bars for Sc/N, B, and C at every reported (L,T,g), and preferably a demonstration","section":"Sign problem, Fig. 2(b)–(f), and SM “Additional data”"},{"comment":"The claim of a first-order transition is based on bimodal histograms at a single size (L=12) and a single temperature (T/t1=1/10), a sharp energy drop in SM Fig. S4(b), and the rapid rise of Sc/N. These are suggestive but not conclusive. No Binder cumulant or histogram analysis is shown for more than one L; no finite-size extrapolation of the energy discontinuity is provided; no hysteresis or metastability is examined. The correlation ratio R_CDW in SM Fig. S3(b) shows a steep crossing, but this is also compatible with a very sharp continuous transition at this size. To support a first-order transition, the authors should show that the bimodal distribution persists and sharpens with increasing L, or provide a quantitative finite-size criterion for the discontinuity (e.g., the L-dependence of the double-peak separation or the extrapolated energy jump).","section":"First-order transition evidence: Fig. 2(a), (d–f); SM Figs. S3–S4"},{"comment":"The Bott index and local Chern marker are computed by identifying the equal-time Green's function with the projector, P≈G. This is an uncontrolled approximation for an interacting, finite-temperature system; G is not a projector and can lose quantization even when the phase is topological. The paper cites Ref. [53] and acknowledges reduced quantization near the transition, but the central narrative that 'topology collapses' at the same coupling as CDW appears relies on this proxy. I would like to see a validation of the G-based marker on a small interacting system where exact diagonalization is possible, or a clearer statement of the accuracy of this approximation. Alternatively, a direct interacting topological invariant or a comparison with the spectral gap would strengthen the claim that the marker collapse is not an artifact of the projector approximation.","section":"Eqs. (3)–(4): Green's-function-based topological markers"}],"minor_comments":[{"comment":"No error bars are shown in any figure. Even if the statistical uncertainties are small, they must be reported to assess the reliability of the histograms and the energy drop.","section":"Throughout"},{"comment":"The color scale for the average sign within each histogram bin is not shown. Add a colorbar so the reader can calibrate the reported effect.","section":"Fig. 2(d)–(f)"},{"comment":"Typographical errors in the SM: 'change-density-wave' in Fig. S2 caption, 'electron-photon coupling' in Fig. S3 caption, and 'radio' instead of 'ratio' in Fig. S2 caption. These should be corrected.","section":"SM captions"},{"comment":"The inset is not described in the caption. Specify what quantity is plotted (presumably Sc/N versus 1/L at fixed g) and its legend.","section":"Fig. 2(b) inset"},{"comment":"The comparison with the DMRG critical point of Ref. [16] should be explained more explicitly: the mapping from Holstein to attractive Hubbard and then to repulsive Hubbard via the particle-hole transformation is stated, but the derivation of U_c/t1≈6.8 from the Holstein parameters (g, ω0) should be written out to avoid ambiguity.","section":"Fig. 4 and mapping to DMRG"},{"comment":"The abstract calls the study 'unbiased,' but DQMC with a complex weight is biased if the sign problem is uncontrolled. I suggest softening this wording to 'numerically exact up to statistical errors controlled by the sign' or similar.","section":"Abstract and text"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about sign-problem control lands directly on the central claim. The manuscript has the right ingredients — no free parameters, multiple independent diagnostics, deposited data — but the missing quantitative sign budget and the single-size histogram evidence make the first-order transition and marker collapse not yet fully convincing. I recommend major revision rather than rejection: the requested analyses are feasible within the existing data and would materially strengthen the paper. The paper would be a strong contribution if these issues are resolved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on the Haldane-Holstein paper. The genuinely new piece is the unbiased DQMC treatment with fully dynamical phonons, giving a finite-temperature phase diagram in (g, t2) and in (lambda, omega0), with a first-order CI-CDW transition at g_c/t1 ~ 1.6 for their parameters. They also bring real-space topological markers and a Bott index constructed from the interacting Green's function, plus OBC and spectral data, and they check both limiting regimes (adiabatic mean-field, antiadiabatic ED). That's a solid, coherent package.\n\nWhat I think holds up: the transition is probably first order. The bimodal histogram at L=12, the sharp energy drop, and the correlation ratio crossing all point that way. And the physical picture—CDW acting as a dynamical Semenoff mass—is reasonable and matches the mean-field limit.\n\nThe soft spot is exactly the one flagged: the sign problem at the transition is not quantified. They show a dip in average sign but don't give numbers, and there are no error bars anywhere in the main figures. Since the histograms are reweighted, and the topological markers are built from equal-time G, statistical control matters. At L=15 with T=1/30 the sign must be small; the SM says DQMC is unfeasible for omega0>4. So the claim that the sign is 'mild' needs support: effective sample sizes or sign values at the reported parameters.\n\nThat said, this is not fatal. The diagnostics are diverse and independently motivated; the sign-proxy is an add-on, not the load-bearing wall. The first-order claim rests on the energy drop and histograms, which are suggestive even without error bars, but a referee should ask for error bars and sign numbers, plus a larger-size histogram or a finite-size study of the bimodality.\n\nSerious thinker: yes, it's clearly reasoned. I'd send it to peer review. The paper will be useful to people working on correlated topological materials and e-ph coupling. I'd cite it.\n\nRecommendation: worth engaging, but the revision should report average sign and error bars explicitly.","headline":"Unbiased DQMC phase diagram with dynamical phonons: a likely first-order CI-CDW transition, but the sign problem at the transition needs quantitative reporting before I'd swallow the details.","tokens_in":18086,"tokens_out":1975,"would_cite":true,"duration_ms":18928,"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":"This paper shows that in the Haldane–Holstein model at half-filling, increasing the electron–phonon coupling drives a first-order transition from a Chern insulator to a staggered charge-density wave, so that the many-body Bott index, the lo","keywords":["Haldane-Holstein model","Chern insulator","charge density wave","Semenoff mass","determinant quantum Monte Carlo","Bott index","local Chern marker","first-order phase transition"],"falsifier":"Compute the average sign and reweighted charge-structure-factor histograms at L = 15, T/t1 = 1/30, and g/t1 ≈ 1.6 with an order of magnitude more Monte Carlo sweeps; if the bimodal distribution or the abrupt energy drop disappears, or if the typical sign falls below roughly 0.1, the first-order CI–CDW transition claim is not established. A complementary check is to compare the Bott-index collapse with exact diagonalization or density-matrix-renormalization-group results on the same parameters, especially in the antiadiabatic limit.","tokens_in":17151,"feed_emoji":"⚛️","tokens_out":6583,"duration_ms":56986,"temperature":0.7,"pith_summary":"The paper asks how retarded electron–phonon interactions destabilize a Chern insulator. Using determinant quantum Monte Carlo, it finds that the Chern insulating ground state survives weak coupling but switches abruptly, through a first-order transition, into a staggered charge-density wave as the coupling g grows. The CDW acts as a spontaneously generated sublattice (Semenoff) mass, which closes and reopens the single-particle gap and removes the chiral edge states exactly where the topological markers lose their quantization. If correct, this establishes a concrete mechanism by which phonons can destroy Chern topology discontinuously, with measurable signatures in spectral and tunneling probes.","feed_headline":"1.6 is where Chern topology gives way to charge order","feed_subtitle":"Quantum Monte Carlo shows a first-order switch to a staggered charge-density wave as phonon coupling grows.","key_machinery":"The argument is carried by three diagnostics used together: the many-body Bott index, the real-space local Chern marker built from the DQMC equal-time Green's function treated as an effective projector, and the staggered charge structure factor Sc. The topological markers establish which phase is a Chern insulator; Sc establishes when charge order is extensive. The conceptual link is the identification of the CDW with a dynamically generated Semenoff mass—a sublattice-staggered potential that breaks sublattice symmetry and thereby destroys the Haldane topology.","core_discovery":"The central claim is that the half-filled Haldane–Holstein model has two competing ground states—a Chern insulator at weak electron–phonon coupling and a staggered charge-density wave at strong coupling—separated by a first-order transition. For t2/t1 = 0.2 and ω0/t1 = 1 the transition sits at g/t1 ≈ 1.6. The evidence presented is simultaneous: the many-body Bott index and the real-space local Chern marker computed from the interacting Green's function remain nearly quantized below the transition and collapse above it; the staggered charge structure factor grows abruptly and becomes extensive; reweighted histograms of that structure factor are bimodal at the transition; the total energy drop","pith_inferences":["The largest system size (L = 15) is exactly where the average sign is lowest, so the most pronounced first-order signatures are also the least statistically controlled; a higher-statistics calculation at that size, or a comparison with a sign-problem-free method at the same parameters, would settle whether the jump is an artifact.","The Green's-function-based Chern marker is known to lose quantization near interacting topological transitions; the conclusion that the transition is topological therefore leans on the marker's reliability, and an independent many-body invariant would harden the claim.","If the mechanism is generic—any interaction that spontaneously generates a sublattice-staggered potential destroys a Chern insulator—then similar first-order CI-to-CDW transitions should appear in other honeycomb and moiré systems with strong electron-phonon coupling; this is testable in candidate TMD heterobilayers.","A concrete experimental fingerprint follows from the paper's own spectral and open-boundary results: at the transition, the edge-state LDOS and the bulk gap should vanish together while the sublattice charge imbalance jumps discontinuously, which could be seen by STM and angle-resolved photoemission in the same sample."],"forward_implications":["At t2/t1 = 0.2 and ω0/t1 = 1, the Chern insulator is stable below g/t1 ≈ 1.6 and the staggered CDW above it, with no coexistence away from the transition point.","The transition is first order, evidenced by bimodal histograms, a sharp total-energy drop, and an energy-level crossing in the antiadiabatic limit.","The average determinant sign dips sharply at the CI–CDW boundary, so the sign itself can serve as a phase-boundary proxy across the g–t2 plane.","Increasing the phonon frequency shifts the critical coupling to larger g; in the antiadiabatic limit the CDW coexists with s-wave superconductivity, and in the adiabatic limit the mean-field critical λ ≈ 2 becomes exact.","Open-boundary simulations show chiral edge states below the transition and their disappearance above it, consistent with gap closing and reopening at the transition."],"fun_headline_variants":["Phonons flip Chern insulator to charge order at g=1.6","First-order switch: Chern topology to charge density wave","Quantum Monte Carlo sees abrupt Chern-to-charge-order transition","Chern topology collapses at a critical electron-phonon coupling","Staggered charge order wins: Chern insulator falls at g=1.6"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that determinant quantum Monte Carlo remains statistically unbiased exactly where the average sign is smallest, at the transition; the paper's own data show a sign minimum there and do not report numerical sign values for the largest lattice, so if the sign problem is severe the first-order jump and reweighted histograms could be numerical artifacts.","fun_headline_variants_meta":{"raw":{"variants":["Phonons flip Chern insulator to charge order at g=1.6","First-order switch: Chern topology to charge density wave","Quantum Monte Carlo sees abrupt Chern-to-charge-order transition","Chern topology collapses at a critical electron-phonon coupling","Staggered charge order wins: Chern insulator falls at g=1.6"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000682,"raw_usage":{"total_tokens":2925,"prompt_tokens":730,"completion_tokens":2195,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":474,"completion_tokens_details":{"reasoning_tokens":2105}},"tokens_in":474,"tokens_out":2195,"duration_ms":14017,"temperature":1.0,"reasoning_tokens":2105,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T02:58:02.800921+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the average sign and reweighted charge-structure-factor histograms at L = 15, T/t1 = 1/30, and g/t1 ≈ 1.6 with an order of magnitude more Monte Carlo sweeps; if the bimodal distribution or the abrupt energy drop disappears, or if the typical sign falls below roughly 0.1, the first-order CI–CDW transition claim is not established. A complementary check is to compare the Bott-index collapse with exact diagonalization or density-matrix-renormalization-group results on the same parameters, especially in the antiadiabatic limit.","supporting_citations":[],"review_version":1}