{"id":"d28b295a-9270-4132-833f-dcc600a1b9cc","arxiv_id":"2607.01682","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"DMC calculations on bilayer InSe show stacking energy differences up to 60 meV per formula unit that depend on full atomic registry and many-body effects, far larger than the near-degeneracy found by DFT.","lead":"This paper benchmarks stacking energies in bilayer InSe with diffusion quantum Monte Carlo and finds much larger energy differences between registries than density functional theory predicts. A smart generalist might read it to see how common computational tools can miss key energy scales in twisted 2D materials used for electronics.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"DMC separation of 8(5) meV/f.u. between AB and AAr lies within ~1.6σ of zero, so the claim that DMC robustly separates all three stackings rests on a statistically marginal difference.","rationale":"The statistical marginality of the 8(5) meV difference is a concrete, quantifiable instance of the convergence-error concern already flagged by the reader. Because the abstract supplies the error bars, this issue is visible even without the full manuscript and directly limits how strongly the reported separations can be used to argue for registry-dependent many-body physics.","tokens_in":1784,"tokens_out":406,"duration_ms":27003,"concrete_test":"Increase DMC walker statistics on the AB and AAr configurations until the stochastic error on their energy difference falls below 2 meV/f.u.; if the central value remains <4 meV or changes sign, the separation between these two stackings is not robust.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that DMC energies for AB, AAr and ABr are distinctly separated (8(5) and 41(4) meV/f.u.) and that the 60(7) meV range demonstrates registry-dependent many-body effects beyond the shared interfacial Se motif. The quoted 8(5) meV difference is only 1.6 standard deviations from zero; at conventional 2σ significance the AB–AAr pair remains consistent with the near-degeneracy reported by DFT (≤1.5 meV). While the 41(4) and 60(7) values are clearer, the headline assertion that “DMC separates these stackings” and that energetics “are not determined by the interfacial atomic motif alone” depends on all three differences being resolved. The reader’s weakest assumption (convergence errors specific to the InSe bilayer) therefore directly threatens the quantitative support for the many-body-response conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper claims that DFT finds AB, AAr, and ABr stackings in bilayer InSe nearly degenerate (within 1.5 meV/f.u.) with similar charge densities, while DMC calculations separate them by 8(5) and 41(4) meV/f.u. with a maximum range of 60(7) meV/f.u.; this demonstrates that local stacking energetics depend on the full registry and many-body electronic response rather than the shared interfacial Se motif alone, implying that DFT-based moiré models substantially underestimate corrugation and its consequences for relaxation and reconstruction.","tokens_in":2000,"tokens_out":415,"duration_ms":16415,"significance":"If the DMC energy differences hold after addressing statistical and convergence issues, the result provides a valuable many-body benchmark for twisted bilayer energetics in InSe, highlighting limitations of DFT local-registry approximations with direct implications for moiré modeling in layered materials. The direct DMC vs. DFT comparison on identical configurations is a strength.","major_comments":[{"comment":"Abstract and results: The reported DMC energy difference of 8(5) meV/f.u. between AB and AAr stackings is only ~1.6σ from zero. This marginal separation undermines the central claim that DMC 'separates these stackings' and that energetics 'are not determined by the interfacial atomic motif alone,' since the headline assertion requires all three registries to be distinctly resolved beyond the DFT near-degeneracy.","section":"Abstract"},{"comment":"Methods/results: No convergence tests, system-size scaling, or detailed methodology (e.g., twist-angle handling, finite-size corrections, or DMC parameters) are provided for the InSe bilayer. This directly threatens verification of the many-body response conclusion and the weakest assumption that the discrete high-symmetry configurations capture the dominant effects without specific convergence errors.","section":"Methods"}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for their careful reading of the manuscript and for highlighting these important points regarding statistical significance and methodological details. We respond to each major comment below.","responses":[{"response":"We agree that the 8(5) meV/f.u. difference is only ~1.6σ and is therefore marginal, which weakens the assertion of clear separation specifically between AB and AAr. The larger 41(4) meV/f.u. separation and 60(7) meV/f.u. overall range still demonstrate that the three stackings are not all near-degenerate within the DFT value of 1.5 meV/f.u., supporting the broader conclusion that energetics depend on the full registry. We will revise the abstract and main text to qualify the AB-AAr result explicitly in terms of its statistical significance while retaining the claim for the resolved separations.","revision_made":"partial","referee_comment":"[Abstract] Abstract and results: The reported DMC energy difference of 8(5) meV/f.u. between AB and AAr stackings is only ~1.6σ from zero. This marginal separation undermines the central claim that DMC 'separates these stackings' and that energetics 'are not determined by the interfacial atomic motif alone,' since the headline assertion requires all three registries to be distinctly resolved beyond the DFT near-degeneracy."},{"response":"The submitted manuscript indeed omits explicit convergence tests, system-size scaling, finite-size corrections, and full DMC parameter details. We will add these in a revised methods section, including justification that the high-symmetry commensurate configurations are appropriate for isolating local-registry effects. Since the study addresses fixed bilayer stackings rather than incommensurate twisted structures, twist-angle handling does not apply and will be clarified as such.","revision_made":"yes","referee_comment":"[Methods] Methods/results: No convergence tests, system-size scaling, or detailed methodology (e.g., twist-angle handling, finite-size corrections, or DMC parameters) are provided for the InSe bilayer. This directly threatens verification of the many-body response conclusion and the weakest assumption that the discrete high-symmetry configurations capture the dominant effects without specific convergence errors."}],"tokens_in":1390,"tokens_out":482,"duration_ms":28669,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The one thing to know is that this paper reports DMC energies separating AB, AAr, and ABr stackings in bilayer InSe by 8(5), 41(4), and 60(7) meV per formula unit, numbers that were not previously quantified. DFT sees them as nearly degenerate within 1.5 meV, so the work shows a clear quantitative gap between the two methods on the same configurations.\n\nWhat the paper does well is run independent DMC and DFT on the same high-symmetry registries and extract the energy differences directly. The abstract states that these stackings share the interfacial Se motif yet still differ under DMC, which supports the point that full registry and many-body response matter beyond the local atomic arrangement. That is a useful concrete result for anyone building moiré models from local DFT energies.\n\nThe soft spot is the 8(5) meV AB–AAr difference. It sits at roughly 1.6 sigma, so it remains consistent with zero at the conventional 2 sigma level. The headline that DMC separates all three registries and that energetics are not set by the interfacial motif alone therefore rests more heavily on the 41(4) and 60(7) values. The abstract also gives no convergence tests, system-size checks, or finite-size error estimates, which leaves the many-body correction claim open to the exact concern raised in the stress test. If those details are solid in the full text the numbers become more reliable; otherwise the smallest separation stays marginal.\n\nThis is for people who model twisted 2D bilayers and need to know how much local corrugation DFT might miss. A reader who works on InSe or similar chalcogenides would get direct numbers to compare against their own calculations. It deserves a serious referee because the question is relevant to the subfield and the method is appropriate, even though the statistical support for the smallest difference will need closer examination.","headline":"DMC gives new stacking energies for InSe that are larger than DFT, but the smallest difference is only 1.6 sigma from zero so the separation claim is not fully robust.","tokens_in":2512,"tokens_out":477,"would_cite":false,"duration_ms":28434,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"DMC shows InSe bilayer stackings differ by up to 60 meV per formula unit while DFT finds them nearly degenerate.","keywords":["bilayer InSe","stacking energetics","diffusion Monte Carlo","DFT benchmarking","twisted bilayers","many-body effects","moiré potentials","van der Waals materials"],"falsifier":"A converged DMC or higher-level calculation on larger supercells that brings the registry energy spread below roughly 10 meV per formula unit, or an experimental probe that measures stacking-energy differences near the DFT values, would falsify the central claim.","tokens_in":2689,"feed_emoji":"⚛️","tokens_out":738,"duration_ms":32041,"temperature":0.7,"pith_summary":"The paper benchmarks local registry energies in bilayer InSe by comparing diffusion quantum Monte Carlo results against density functional theory. DMC finds that three stackings sharing the same interfacial Se atoms are separated by 8 and 41 meV per formula unit, with the full range reaching 60 meV, whereas DFT keeps them within 1.5 meV. The separation demonstrates that registry energy is set by the complete atomic arrangement and its many-body electronic response, not by the interface motif alone. This directly affects how accurately DFT-based moiré models can describe relaxation, domains, and electronic structure in twisted bilayers.","feed_headline":"DMC reveals 60 meV registry energy spread in InSe bilayers","feed_subtitle":"DFT keeps three stackings within 1.5 meV while many-body calculations separate them by tens of meV, showing interface motif is not enough.","key_machinery":"Diffusion quantum Monte Carlo (DMC) applied to the three high-symmetry stackings of bilayer InSe to extract registry-dependent total energies beyond DFT.","core_discovery":"DMC separates AB, AAr, and ABr stackings by 8(5) and 41(4) meV/f.u., while the energy difference between the most stable and least stable registries reaches 60(7) meV/f.u.. These large energy separations show that the stacking energetics are not determined by the interfacial atomic motif alone but depend on the full registry and its associated many-body electronic response. More broadly, these results show that DFT-based moiré models can substantially underestimate local stacking-energy corrugation, with direct consequences for predicted structural relaxation, domain formation, and electronic reconstruction in twisted layered materials.","pith_inferences":["Models that fit moiré potentials from DFT alone may need systematic corrections derived from many-body methods for quantitative predictions of twist-angle-dependent properties.","Local stacking energies could be measured by combining atomic-resolution imaging with spectroscopy to test the DMC scale directly.","The many-body contribution may also affect phonon modes or excitonic binding that depend on interlayer registry."],"forward_implications":["Twisted InSe bilayers will exhibit stronger structural relaxation and larger domain sizes than DFT moiré models predict.","Electronic reconstruction at domain walls will be more pronounced because of the deeper local energy wells.","Similar underestimation of stacking corrugation is expected in other van der Waals bilayers when only DFT is used."],"fun_headline_variants":["InSe registries span 60 meV under DMC","DMC separates InSe stackings by 60 meV","Bilayer InSe has 60 meV registry energy range","Many-body calc yields 60 meV InSe energy gaps"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The three chosen high-symmetry stackings represent the dominant local registries that appear in a continuously twisted bilayer and that the DMC energies for these discrete cells capture the main many-body contributions without large finite-size or convergence errors.","fun_headline_variants_meta":{"raw":{"variants":["InSe registries span 60 meV under DMC","DMC separates InSe stackings by 60 meV","Bilayer InSe has 60 meV registry energy range","Many-body calc yields 60 meV InSe energy gaps"]},"model":"grok-4.3","cost_usd":0.006862,"raw_usage":{"total_tokens":3204,"prompt_tokens":703,"num_sources_used":0,"completion_tokens":67,"cost_in_usd_ticks":68624500,"prompt_tokens_details":{"text_tokens":703,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2434,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":703,"tokens_out":67,"duration_ms":33126,"temperature":1.0,"reasoning_tokens":2434,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-03T10:19:24.454607+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A converged DMC or higher-level calculation on larger supercells that brings the registry energy spread below roughly 10 meV per formula unit, or an experimental probe that measures stacking-energy differences near the DFT values, would falsify the central claim.","supporting_citations":[],"review_version":1}