{"id":"69651c3b-ebff-4eb2-8af9-7fa7225625f4","arxiv_id":"2509.12311","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In the SU(N) two-channel Kondo lattice on a square lattice, exact simulations show antiferromagnetic order at weak coupling and a spontaneous (pi,0)/(0,pi) stripe channel-order phase at strong coupling for N>=6.","lead":"This paper uses exact quantum Monte Carlo to map the phase diagram of a two-channel Kondo lattice in two dimensions, finding antiferromagnetism at weak coupling and a stripe-shaped channel order at strong coupling for N>=6. It matters because it provides unbiased numerical evidence on an open question about channel symmetry breaking in correlated heavy-electron models.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Stripe order extrapolation relies on unverified exclusion of L=4n sizes; if the shifted peaks are not finite-size artifacts, the central phase may not survive.","rationale":"The paper is methodologically strong in many respects: sign-problem-free DQMC, clear benchmarks against exact diagonalization, careful Trotter and Uf checks, and a large-N analysis that provides a plausible mechanism for stripe order in an intermediate coupling window. However, the central claim of a stripe channel symmetry-broken phase rests on the finite-size extrapolation of the dimer order parameter. The Supplemental Materials explicitly exclude all L=4n sizes because the Bragg peak shifts to neighboring momenta, and the assertion that both parity sequences converge to the same (0,pi) ordering is not backed by shown data. This is not a fundamental flaw but a testable gap: the authors should demonstrate that the excluded sequence also extrapolates to a nonzero d((0,pi)) or provide a physical reason why the L=4n+2 shell-filling sequence is the correct one. A secondary inconsistency is that the large-N analysis predicts ferrochannel order for J/t>3.23, while the abstract claims stripe order at 'sufficiently strong Kondo coupling'; this does not overturn the numerical observation but weakens the 'support' claim and should be reconciled. Both issues are addressable, so conditional acceptance is appropriate rather than outright rejection.","tokens_in":12809,"tokens_out":6745,"duration_ms":64266,"concrete_test":"Run DQMC for N=8 at J/t=1.6 (in the claimed stripe phase) for the excluded sizes L=8,12,16 and, if feasible, L=20. For each L, locate the momentum k_max(L) that maximizes D(k,L) and also evaluate D at k=(0,pi). Extrapolate both k_max(L) and D(k_max,L) to L→∞ using all L (or separately for each parity sequence). If k_max(L)→(0,pi) and the extrapolated d((0,pi)) agrees with the L=4n+2 result within error bars, the exclusion is justified. If k_max(L) converges to a wave vector different from (0,pi), or if the order parameter extrapolates to zero when the 4n sizes are included, the stripe phase is a finite-size artifact.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central QMC evidence for the stripe phase is the dimer order parameter d(k=(pi,0)) extrapolated to L→∞. In the Supplemental Materials, the authors state: 'To accelerate convergence, we exclude system sizes L=4,8,12,16, relying solely on L=6,10,14,18.' They justify this by observing that for L=4n the maximum in D(k,L) shifts to 'the nearest available momenta around these stripe wave vectors,' and assert that both sequences converge in the thermodynamic limit to the same (0,pi) stripe ordering. However, no data or extrapolation for the L=4n sequence is provided. Because the L=4n+2 sequence is commensurate with a (0,pi) stripe (since the stripe period 2 divides L), the chosen sizes can artificially lock in the commensurate order. If the true thermodynamic state has an incommensurate wave vector that approaches but never equals (0,pi), the L=4n+2 extrapolation would overestimate d((0,pi)) and produce a spurious nonzero order parameter. The existence of the stripe phase therefore hinges on an unverified assumption about which sequence is correct.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the SU(N)-symmetric two-channel Kondo lattice model on the square lattice at half-filling using sign-problem-free determinant quantum Monte Carlo for N = 2, 4, 6, 8, for system sizes up to L = 18. The authors report antiferromagnetic order for weak Kondo coupling for all N, and, for N >= 6 at sufficiently strong coupling, a channel-symmetry-broken phase characterized by a (pi,0)/(0,pi) stripe dimerization pattern. The numerical results are complemented by a large-N saddle-point analysis comparing ferrochannel, antiferrochannel, and pi-modulated stripe hybridization ansatzes, and by a Fermi-surface reconstruction analysis.","tokens_in":13162,"tokens_out":2773,"duration_ms":33416,"significance":"If the central claim holds, this is a valuable result: a two-dimensional unbiased numerical demonstration of spontaneous channel symmetry breaking with stripe order in a Kondo lattice, going beyond both large-N and one-dimensional studies. The methodological strengths are substantial: the DQMC approach is sign-problem-free for the model considered, the implementation is benchmarked against exact diagonalization in the single-site limit, and the large-N calculation is parameter-free in the sense that it derives the order parameter from the saddle point. However, the thermodynamic-limit inference for the stripe phase relies on a finite-size-selection assumption that is not fully supported, and the large-N 'corroboration' is partly constructed after the numerical observation. These issues are load-bearing for the paper's main claim.","major_comments":[{"comment":"The paper excludes system sizes L = 4, 8, 12, 16 from the thermodynamic-limit extrapolation and uses only L = 6, 10, 14, 18, stating that both sequences 'converge in the thermodynamic limit to the same (0,pi) stripe ordering,' but no data or extrapolation for the L = 4n sequence is shown. This is a load-bearing assumption: if the shifted Bragg peaks seen at L = 4n are not finite-size artifacts, the stripe order parameter d((pi,0)) could be overestimated or the true ordering wave vector could be incommensurate. The authors must either provide the L = 4n sequence data and a clear demonstration of convergence to the same (0,pi) ordering, or otherwise rigorously justify the exclusion (e.g., by a finite-size scaling of the peak position and amplitude for both sequences). Without this, the central claim of stripe long-range order is not established.","section":"Supplemental Materials, 'Spatial Structure of the Channel Symmetry Breaking' and Fig. S3"},{"comment":"The large-N calculation is not an independent test of the stripe hypothesis because the ansatz space was chosen based on the QMC results: the text states 'Motivated by our numerical simulations in the main text, we compare the ground-state energies of three specific spatial patterns.' The subsequent agreement is therefore a retrospective fit rather than a falsifiable prediction. This should be stated explicitly and the corroborative value adjusted. In addition, the large-N analysis finds the stripe ansatz lower in energy only for J/t <= 3.23, while the QMC stripe signal is reported at stronger couplings; the ferrochannel state preferred at large J in the large-N limit is 'not observed in our numerical results.' This discrepancy, acknowledged in the text, further weakens the claimed analytical support and should be addressed or explicitly left as an open question.","section":"Supplemental Materials, 'Large-N analysis'"},{"comment":"The finite-size scaling uses the ansatz O(L) = O_inf + A1/L^2 + A2/L^4 on a hand-selected sequence of four sizes (6, 10, 14, 18), sometimes omitting the L=6 point or setting A2=0. With only 4 points and up to 3 parameters, the quality of the fit and the stability of O_inf under alternative (e.g., L=4n) sequences, different fit ranges, or different correction exponents should be reported. The claim that L=4n+2 sizes provide a 'nice sequence' is plausible but not quantitatively justified; the sensitivity of the extracted order parameter to these choices affects the phase boundaries and the coexistence claim for N=8.","section":"Eq. (S16) and Fig. S3/S4"}],"minor_comments":[{"comment":"The text 'diff' appears truncated in the caption: 'stripe channelπ t t J1 J2 c1 c2 f' and 'diff' likely incomplete; please fix.","section":"Fig. 2 caption"},{"comment":"The statement 'we verified that this ensures that the half-filling constraint in Eq. (2) holds' is supported by Fig. S7, but the text could quote the actual variance values for clarity, especially at larger N where the constraint is less tight.","section":"Main text, paragraph after Eq. (3)"},{"comment":"The definition of D(k,L) for kx != ky as the sum of C4-related configurations should specify whether this sum is normalized and how the 'nearest available momenta' are defined in the supplemental discussion; a reader cannot reproduce the data without this detail.","section":"Main text, 'Observables'"},{"comment":"The derivation of the order parameter from the derivative with respect to Delta J_a should mention that the derivative is taken at fixed V*; the notation is slightly ambiguous.","section":"Supplemental Materials, Eq. (S15)"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a relevant problem in strongly correlated electron systems and uses a technically sound method. The main concern is not the methodology per se but the selective reporting of finite-size data in the supplemental materials: the exclusion of L=4n sizes is essential to the stripe claim, yet no data for that sequence is shown. This is unusual for a central phase-diagram claim and should be addressed head-on. The large-N 'support' is also less compelling than it initially appears, because the ansatz space was informed by the numerics. I would not reject the paper: the QMC results are sign-problem-free and benchmarked, and the observation of a (pi,0) dimer peak is suggestive. But the thermodynamic-limit conclusion needs a more transparent and more robust finite-size analysis before publication. The fit to the journal's scope is appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a technically strong paper that delivers the first sign-problem-free determinant QMC study of the two-channel Kondo lattice in 2D, and it reports a genuinely new phase—a pi-modulated stripe channel order for N>=6. If the stripe order survives a closer look, it's a substantial result. The right response is to referee it, but with a specific demand: show the excluded L=4n data.\n\nWhat's new and good: The method is sign-problem-free at half-filling for even N, which is a real achievement for this model. The single-site ED benchmark in the supplementary is careful. The AFM to stripe transition as a function of J for N=6,8 is presented with finite-size extrapolations, and the large-N saddle-point analysis includes the stripe ansatz, which isn't in earlier work. The Fermi surface reconstruction comparison between QMC and large-N is convincing support for the stripe pattern.\n\nSoft spots: The central claim rests on the L=4n+2 sequence (6,10,14,18), with L=4,8,12,16 excluded. The authors say the excluded sizes show peaks at nearest available momenta and that both sequences converge, but they don't show the excluded data. That's a genuine gap; the extrapolation could be biased if the true ordering wave vector is incommensurate and only approaches (0,pi) in the thermodynamic limit. This isn't a fatal flaw—the commensurate sequence may well be the right one—but it needs a direct response. Second, the large-N analysis has a tension: for J/t>3.23 it prefers the ferrochannel state, while their QMC sees stripe at those couplings. They mention 1/N corrections, but the discrepancy should be quantified or at least addressed with data at larger J.\n\nMinor: the AFM order parameter for N=2,4 is measured out to J=8.2 with no sign of channel order; that's fine but worth checking whether the f-spin constraint is still holding at these large J.\n\nBottom line: this deserves a serious referee. The paper is honest, the numerics are sign-problem-free and benchmarked, and the stripe phase is interesting enough to be chased. Ask for the L=4n data and a more careful discussion of the large-N mismatch before acceptance.","headline":"First sign-problem-free DQMC for the 2D two-channel Kondo lattice: a genuinely new stripe channel phase for N>=6, with a real but addressable concern about the excluded L=4n sizes.","tokens_in":13573,"tokens_out":2896,"would_cite":true,"duration_ms":31730,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.27.+a","75.30.Mb"],"model":"deepseek-v4-flash","headline":"In the half-filled SU(N) two-channel Kondo lattice, strong Kondo coupling with N≥6 spontaneously breaks channel symmetry and forms a stripe phase.","keywords":["two-channel Kondo lattice","SU(N) symmetry","antiferromagnetism","channel symmetry breaking","stripe order","determinant quantum Monte Carlo","large-N saddle point","Fermi surface reconstruction"],"falsifier":"Compute the dimer correlation D(k) for L = 22 and L = 26, or reanalyze the excluded sizes L = 4, 8, 12, 16 with a more careful finite-size scaling, and check whether the (0,π) Bragg peak extrapolates to the same nonzero thermodynamic value; if the excluded sequence converges to a different order or to zero, the stripe phase would be a finite-size artifact.","tokens_in":12739,"feed_emoji":"🧲","tokens_out":4814,"duration_ms":52356,"temperature":0.7,"pith_summary":"The paper determines the zero-temperature phase diagram of the SU(N)-symmetric two-channel Kondo lattice model on a square lattice at half filling, using sign-problem-free determinant quantum Monte Carlo for N = 2, 4, 6, and 8. It finds that at weak Kondo coupling the localized moments order antiferromagnetically for all N. For N ≥ 6, increasing the Kondo coupling destroys the antiferromagnet and induces spontaneous channel-symmetry breaking, with the hybridization alternating between the two conduction channels in a stripe pattern at wave vector (π,0)/(0,π). The quantum Monte Carlo results are corroborated by a large-N saddle-point analysis that finds the striped hybridization pattern lowest in free energy at low and intermediate coupling. If correct, this establishes unbiased numerical evidence for a stripe-like channel-ordered phase in a two-dimensional two-channel Kondo lattice.","feed_headline":"Kondo lattice orders into stripes at strong coupling","feed_subtitle":"Sign-problem-free simulations and large-N analysis locate a stripe phase that reconstructs the Fermi surface.","key_machinery":"The central object is the channel-magnetization (dimer) operator M_i^z, which measures the difference in singlet formation between the two conduction channels; its correlation function D(k) at k = (π,0) detects stripe order. On the analytic side, Hubbard-Stratonovich decoupling introduces a complex hybridization field V_{ia}, and comparing static ferrochannel, antiferrochannel, and π-modulated stripe ansätze in the large-N limit identifies the stripe configuration as the free-energy minimum at low and intermediate coupling. The simulations use sign-problem-free determinant quantum Monte Carlo at half filling for even N, combined with finite-size extrapolation along the L = 4n+2 sequence.","core_discovery":"At half filling on the square lattice, the ground state evolves from a Kondo-screened antiferromagnet at small J/t into, for N ≥ 6, a channel-symmetry-broken phase in which one channel preferentially hybridizes with localized moments in a stripe pattern with ordering wave vector (0,π)/(π,0). The stripe order parameter d(k) becomes nonzero above J_c ≈ 1.8t for N = 6 and ≈ 1.3t for N = 8, and the reconstructed Fermi surface shows one-dimensional metallic channels. The stripe order is robust to small explicit channel asymmetry (destabilized only at ΔJ/J ≈ 0.025 for N = 6), and for N = 8 antiferromagnetic and stripe order coexist in the numerically accessible regime.","pith_inferences":["We infer that the stripe phase is a strong-coupling counterpart of the RKKY antiferromagnet: channel symmetry breaking absorbs the Kondo energy while the stripe pattern relieves the intersite magnetic competition, and the two phases meet near J_c.","If the stripe order survives the thermodynamic limit, the accompanying Goldstone-like meandering of stripe domain walls would realize an electronic smectic in a microscopic model, linking this phase to stripe physics proposed for high-temperature superconductors.","The discrepancy between the numerical preference for stripes at large J and the large-N prediction of ferrochannel order above J/t ≈ 3.23 suggests that 1/N corrections or channel fluctuations favor stripes; extending the simulations to larger N or to 1/N-improved saddle points could settle this.","Away from half filling, where the sign problem returns, the stability of stripe order under doping remains open; numerical methods without a sign problem, or analytic continuation approaches, could test the phase's robustness."],"forward_implications":["For N ≥ 6 at strong Kondo coupling, the ground state is a stripe channel-ordered metal whose Fermi surface reconstructs into one-dimensional-like bands.","The stripe phase persists under small explicit channel asymmetry, indicating it is not merely a fine-tuned consequence of exact SU(2) channel symmetry.","The large-N saddle-point analysis predicts a crossover to uniform ferrochannel order at J/t > 3.23, a competition that the numerics do not yet resolve.","For N = 2 and 4, no channel symmetry breaking is found in the studied coupling range; antiferromagnetic order persists up to J/t = 8.2.","The phase diagram provides a concrete benchmark for two-channel Kondo lattice physics in two dimensions, complementing earlier one-dimensional and infinite-dimensional studies."],"fun_headline_variants":["Kondo lattice stripes appear for N≥6 at strong coupling","Striped channel order emerges in SU(N) Kondo lattice","Strong coupling drives stripe order in two-channel Kondo","N≥6 Kondo lattice shows striped hybridization","Kondo stripe order reconstructs Fermi surface"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The thermodynamic-limit stripe order rests on the assumption that only the L = 6, 10, 14, 18 system sizes provide the correct finite-size sequence, and that the shifted dimer Bragg peaks seen at L = 4, 8, 12, 16 are finite-size artifacts.","fun_headline_variants_meta":{"raw":{"variants":["Kondo lattice stripes appear for N≥6 at strong coupling","Striped channel order emerges in SU(N) Kondo lattice","Strong coupling drives stripe order in two-channel Kondo","N≥6 Kondo lattice shows striped hybridization","Kondo stripe order reconstructs Fermi surface"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00061,"raw_usage":{"total_tokens":2662,"prompt_tokens":713,"completion_tokens":1949,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":457,"completion_tokens_details":{"reasoning_tokens":1872}},"tokens_in":457,"tokens_out":1949,"duration_ms":15455,"temperature":1.0,"reasoning_tokens":1872,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T16:36:15.610756+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the dimer correlation D(k) for L = 22 and L = 26, or reanalyze the excluded sizes L = 4, 8, 12, 16 with a more careful finite-size scaling, and check whether the (0,π) Bragg peak extrapolates to the same nonzero thermodynamic value; if the excluded sequence converges to a different order or to zero, the stripe phase would be a finite-size artifact.","supporting_citations":[],"review_version":1}