{"id":"f62c1e8a-8188-4ccd-b911-0d5eb63a4e3d","arxiv_id":"2608.07372","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":8,"one_line_summary":"A microscopic 2D model with trivial conduction and valence bands is shown to host a Chern exciton band with C=+1, generated purely by interactions.","lead":"Physicists constructed a 2D model where the electrons and holes come from ordinary, topologically trivial bands, yet the bound electron-hole pairs (excitons) form a topological band with nonzero Chern number. This shows that interactions alone can create topological excitons, providing a recipe for designing such states in real materials.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The five-state truncation of the relative-coordinate basis is the least secure step: the Chern number is computed in a 5×5 and then 441×441 projected space, but no proof or convergence test establishes that the Wilson-loop winding survives the infinite-Δ limit.","rationale":"The reader's weakest assumption is exactly the one I find most load-bearing: the topological invariant is established in a truncated relative-coordinate basis, and the infinite-Δ limit is not rigorously controlled. The paper does provide a meaningful partial check by preserving the gap in a 441×441 basis, and the Wilson loop in that enlarged basis shows the expected winding, which substantially mitigates the concern. For a theoretical construction paper, this level of numerical evidence is normally sufficient, and the construction is internally consistent: the inversion-eigenvalue analysis, the emergent time-reversal-symmetry obstruction, the symmetry-allowed V′ term that breaks it, and the Wilson-loop and ribbon-edge checks all point in the same direction. I therefore do not see a reason to reject or to make acceptance conditional; the truncation issue is a real caveat but not a demonstrated flaw. A cutoff-convergence study would settle it definitively, but its absence does not overturn the existence proof as presented.","tokens_in":22535,"tokens_out":15974,"duration_ms":154613,"concrete_test":"Recompute the lowest-band Wilson-loop winding and the minimum gap above it for cutoffs L = 10, 20, 40, and 80 (|Δx|,|Δy| ≤ L) at the parameters of Fig. 2, and also fit the decay of |φ_{n,Δ}(p)| at Γ and M. If the winding stays +1 and the gap remains open for all L, the truncation concern is settled; if the winding changes at any L, the central claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is an existence proof for a Chern exciton in the full projected exciton Hamiltonian, yet the topological invariant is obtained after truncating relative displacements to Δ ∈ {0, ±x̂, ±ŷ}, and subsequently to |Δx|,|Δy| ≤ 10. The five-state basis is where the inversion-eigenvalue pattern (+,+,+,-) and the band inversion are engineered, while the Wilson loop of Fig. 2(c) is computed in the enlarged 441×441 basis. What would have to be true for the claim to hold is that the winding of the lowest exciton band is unchanged as the basis is enlarged to infinity. The paper checks that the bulk gap above the lowest band is preserved in the 441×441 basis, but it does not provide a sequence of cutoffs or a decay bound establishing that the Wilson-loop winding is converged. The Supplemental Material (Sec. S2 A4) explicitly says the truncation is justified 'provided that the additional terms introduced beyond the atomic limit are sufficiently small not to close this energy separation,' but no quantitative bound is given for the chosen parameters. Since the central claim concerns the actual model, not a finite truncation of it, a distant-envelope state that inverts with the lowest band at some larger Δ would change C_exc and invalidate the existence proof.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs an explicit two-dimensional lattice model in which the electron and hole bands are topologically trivial, yet the lowest exciton band carries Chern number C_exc = +1 and the second band C_exc = -1. The construction works in a real-space exciton basis parameterized by relative electron-hole displacement Δ, starting from the atomic limit of a two-sublattice insulator. A sequence of density-density interactions, a pair-hopping interaction, and weak hoppings induces a band inversion in the five-state subspace Δ ∈ {0, ±x̂, ±ŷ}. The paper proves an emergent spinless time-reversal symmetry of the projected Hamiltonian for all inversion-preserving hoppings and density-density interactions, and shows that a correlated interaction term is required to break this symmetry. The nontrivial topology is diagnosed by inversion eigenvalues and confirmed by Wilson-loop winding and chiral edge states in a ribbon. The paper also presents an oblique-lattice realization and a route to multiband generalizations.","tokens_in":22762,"tokens_out":8277,"duration_ms":68517,"significance":"The central claim—an explicit microscopic existence proof for interaction-induced Chern excitons in trivial electronic bands—is important and, if correct, would resolve a question left open by the authors' earlier classification work. The paper is technically strong in several respects: the projected Hamiltonian is derived explicitly in the Supplemental Material; the emergent time-reversal obstruction is proven for a broad class of terms; and the topology is checked by three independent probes (inversion eigenvalues, Wilson loops, edge states). The construction recipe is concrete and transferable. The main caveat is the basis truncation, which does not yet have a rigorous or quantitative justification.","major_comments":[{"comment":"The five-state truncation Δ ∈ {0, ±x̂, ±ŷ} is load-bearing for the existence proof. The Chern number is computed for this truncated basis, and the enlarged 441×441 basis is used only to verify that the bulk gap above the lowest band is preserved; no convergence test is provided for the Wilson-loop winding itself. The Supplemental Material (Sec. S2 A4) states that the truncation is justified 'provided that the additional terms introduced beyond the atomic limit are sufficiently small not to close this energy separation,' but no quantitative bound is given for the parameters used in the main text. Since the central claim concerns the full projected exciton Hamiltonian, an unobserved distant-envelope state that inverts with the lowest band at larger |Δ| would change C_exc. I request either a convergence test of the Wilson-loop winding for a sequence of cutoffs (e.g., |Δ_x|,|Δ_y| ≤ 5, 10, 15, 20) or a rigorous decay bound on the low-energy envelope wave functions in Δ that guarantees the winding is unchanged.","section":"Main text, 'Inversion-guided exciton band inversion'; SM Sec. S2 A4"},{"comment":"The paper's headline claim is that the conduction and valence bands are topologically trivial. Although the construction starts from the atomic limit, the subsequent addition of t_x and t_{x+y} hoppings changes the non-interacting electronic Hamiltonian; the paper never explicitly verifies that both bands remain Chern-trivial in the full model. A one-sentence argument using the reality of the hoppings and the resulting spinless time-reversal symmetry would suffice, but without it the claim of interaction-induced topology is not fully closed.","section":"Full model definition, SM Eq. (S27)"}],"minor_comments":[{"comment":"The identity blocks in Eq. (2) are written as '1 1×1' and '1 2×2'; consider using standard notation such as I_1 and I_2 to avoid confusion with the inversion operator I.","section":"Eq. (2)"},{"comment":"The ribbon calculation is described as using the same parameters as the bulk, but the parameters are not restated in the figure caption; please list them for completeness.","section":"Fig. 2(d)"},{"comment":"The inversion eigenvalues of the lowest band are given as (-1,-1,-1,+1) in the SM and as (+,+,+,-) in the main text, with the difference attributed to an absorbed overall sign; this notation shift is confusing and should be flagged explicitly at first use in each context.","section":"SM Sec. S4 B"},{"comment":"The 'Note added' mentions a very recent independent work (Ref. [56]); a brief sentence describing the difference between the two constructions would help the reader position this work.","section":"Introduction, Note added"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely to be accepted after the truncation-convergence question is addressed. The authors have a strong track record and the multiple independent checks are persuasive, but the central claim's validity for the infinite-basis limit should not be taken on faith. I would not recommend rejection; the issue is fixable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, this is the first explicit 2D model where both electronic bands are topologically trivial and the lowest exciton band carries Chern number +1. Second, the evidence is strong: inversion eigenvalues, Wilson-loop winding, and ribbon edge states all agree, and the supplemental material contains the full projected Hamiltonian derivation. It is a legitimate existence proof, not a sketch.\n\nWhat is actually new: previous Chern excitons lived in Chern bands; the authors' own earlier work predicted that trivial bands could host shift excitons in 1D and gave a classification but no 2D construction. This paper builds that construction. The model is simple enough to be followed line by line: an atomic limit with two sublattices, density-density interactions that create a dispersive band plus two doublets, a V_x term that induces the band inversion, and a V' term that breaks the emergent spinless time-reversal symmetry which otherwise protects nodal lines. The proof of that emergent TRS—that all hoppings and density-density interactions preserve it, so you need a non-density interaction to break it—is clean and is the kind of general statement that will outlive this particular model.\n\nThe soft spots, in proportion. The biggest is the truncation of the relative-coordinate basis. The Chern number is computed in a five-state basis and then in a 441-state basis, but the full projected Hamiltonian is infinite-dimensional. The paper does not show that the Wilson-loop winding survives the infinite-Δ limit; the SM justifies the truncation by an energy-separation argument but gives no quantitative bound for the chosen parameters. This is exactly the concern the stress-test note raises, and it is accurate. I do not think it kills the paper: the 441×441 calculation is a genuine convergence check of the low-energy spectrum, the gap is preserved, and for short-range interactions a strongly bound exciton should decouple from the far-separation continuum. But a referee should ask for a sequence of cutoffs or a decay estimate to make the existence proof airtight.\n\nThe other soft spot is that the essential V' interaction is engineered, with no microscopic mechanism proposed. The paper says so explicitly in the discussion. That is an honest limitation, not a hidden one.\n\nWho this is for: anyone working on exciton topology, moiré systems, or interaction-induced topological phases. It is a proof of principle with a clear recipe, not a material proposal. It deserves a serious referee. I would send it out. If I were refereeing, I would ask the authors to address the truncation convergence directly, and to double-check the stable-zero formulas from Ref. 48, but I would not block it on either.","headline":"A concrete, well-checked existence proof for Chern excitons from trivial bands, with a truncation caveat that is real but not disqualifying.","tokens_in":23352,"tokens_out":3457,"would_cite":true,"duration_ms":30369,"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 constructs a 2D model with topologically trivial electron bands whose lowest exciton band carries Chern number +1, showing interactions alone can generate exciton topology.","keywords":["Chern exciton","exciton topology","interaction-induced topology","inversion symmetry","Wilson loop","envelope wave function","real-space exciton basis","trivial bands"],"falsifier":"Enlarge the relative-coordinate cutoff from $|\\Delta_x|,|\\Delta_y|\\le 10$ to systematically larger values and recompute the Wilson-loop winding of the lowest exciton band; if the winding ever departs from $C_{\\mathrm{exc}}=+1$, or the bulk gap closes against the exciton continuum before the infinite-basis limit, the topological claim fails.","tokens_in":22242,"feed_emoji":"⚛️","tokens_out":11515,"duration_ms":90989,"temperature":0.7,"pith_summary":"Excitons are bound electron-hole pairs, and their topology is usually inherited from the topology of the underlying electron bands. This paper asks whether interactions alone can create a topological exciton when both participating bands are ordinary insulators, and answers yes by explicit construction. The authors build a two-dimensional inversion-symmetric model whose conduction and valence bands are topologically trivial, then show that the lowest exciton band carries Chern number $C_{\\mathrm{exc}}=+1$, diagnosed by inversion eigenvalues, Wilson-loop winding, and chiral edge states in a ribbon. The topology lives in the exciton envelope wave function—the way the electron-hole separation is distributed—rather than in the bands. If correct, this is a microscopic existence proof and a recipe for engineering interaction-induced topological excitons.","feed_headline":"Interactions alone can create Chern excitons from trivial bands in 2D","feed_subtitle":"A 2D model with trivial electron bands yields a Chern exciton, showing interactions alone can generate exciton topology.","key_machinery":"The load-bearing object is the projected exciton Hamiltonian $H_{\\Delta,\\Delta'}(\\mathbf{p})$ in the real-space relative-distance basis $|\\Delta,\\mathbf{p}\\rangle$, truncated to the five shortest electron-hole separations $\\Delta\\in\\{0,\\pm\\hat{x},\\pm\\hat{y}\\}$. Here $\\Delta$ is the lattice vector separating the electron and hole and $\\mathbf{p}$ is the total exciton momentum, so the eigenvectors $\\phi_{n,\\Delta}(\\mathbf{p})$ are the exciton envelope wave functions. This finite-dimensional Hamiltonian reduces the exciton-topology question to a conventional band-topology problem on the envelope space, with inversion represented by $I=1_{1\\times1}\\oplus\\sigma_x\\oplus\\sigma_x$; inversion eigenvalues at the four high-symmetry points then diagnose the Chern number modulo two. The second essential piece is the emergent spinless time-reversal symmetry: any projected Hamiltonian built from real hoppings and density-density interactions obeys $H_{\\Delta,\\Delta'}(\\mathbf{p})=(H_{\\Delta,\\Delta'}(-\\mathbf{p}))^*$, which protects nodal points and must be broken by the specifically designed complex interaction $V'$ to expose the Chern topology.","core_discovery":"The paper's central discovery is that a gapped Chern exciton band can be manufactured from two topologically trivial bands through carefully chosen interactions. Starting from the atomic-limit Hamiltonian $H_0=-t_0\\sum_{\\mathbf{R}}(c_{\\mathbf{R},A}^\\dagger c_{\\mathbf{R},B}+\\mathrm{h.c.})$ on a two-sublattice lattice, the valence and conduction Wannier states are the even and odd sublattice combinations, with zero Chern number. Writing the projected exciton Hamiltonian in the relative-distance basis $\\Delta\\in\\{0,\\pm\\hat{x},\\pm\\hat{y}\\}$, the authors add repulsive density-density interactions $U_0,U_x,U_y$, then a symmetry-allowed interaction $V_x$ that induces a band inversion between the envelope states $\\Phi_0$ and $\\Phi_{-x}$ at the $M$ point. Because the projected Hamiltonian inherits a spinless time-reversal symmetry from ordinary hoppings and density-density terms, the inversion alone gives only nodal points; a further inversion-symmetric interaction $V'$ with a factor of $i$ breaks that emergent symmetry and opens a full gap. The resulting lowest exciton band has inversion eigenvalues $(+,+,+,-)$ at $\\Gamma,X,Y,M$, giving $C_{\\mathrm{exc}}=+1$; Wilson loops give $C_{\\mathrm{exc}}=+1$ and $-1$ for the two lowest bands, and a ribbon calculation shows chiral edge states. The Chern number is carried entirely by the exciton envelope wave function, independent of the trivial electronic bands.","pith_inferences":["This suggests that materials with tightly bound excitons and tunable interactions—such as moiré heterostructures or coupled quantum wells—could be screened for chiral exciton edge modes even when their single-particle bands are trivial.","One testable extension is to interpret the five-state envelope Hamiltonian as an effective tight-binding model on the relative-coordinate lattice; the Chern number would then be tunable by engineering envelope hoppings and phases instead of the electron bands.","The emergent spinless time-reversal symmetry may be a general obstruction: in any exciton projection from real hoppings and density-density terms, a nonzero exciton Chern number requires some term that acts on the envelope and breaks this effective symmetry, which narrows the search for realistic microscopic mechanisms."],"forward_implications":["Interaction-induced Chern excitons are possible in principle: no topological parent bands are required, so trivial-band materials are not ruled out as hosts.","The inversion eigenvalues of the lowest exciton band, $(+,+,+,-)$, determine $C_{\\mathrm{exc}}=+1$, while the second band carries $C_{\\mathrm{exc}}=-1$, as confirmed by Wilson-loop winding.","Chiral exciton edge states traverse the bulk gap in a ribbon geometry, giving a concrete transport signature of the interaction-induced topology.","Any projected exciton Hamiltonian built from real hoppings and density-density interactions inherits spinless time-reversal symmetry, so achieving a Chern exciton requires an interaction term outside that class.","The construction recipe—atomic limit, inversion-guided band inversion, and breaking the emergent symmetry—provides a general route for designing topological excitons in other lattice settings."],"supporting_citations":[{"why":"Supplies the real-space exciton basis and projected-Hamiltonian formalism on which the construction is built.","marker":"[43]"},{"why":"Provides the inversion-symmetry classification and the $C_{\\mathrm{exc}}\\bmod 2$ criterion from inversion eigenvalues used to choose the band inversion.","marker":"[48]"},{"why":"Establishes the exciton Wilson-loop method used to compute $C_{\\mathrm{exc}}=+1$ and $-1$ for the two lowest bands.","marker":"[44]"},{"why":"Gives the inversion-eigenvalue formula relating odd-parity eigenvalue counts to Chern number modulo two, extended here to excitons.","marker":"[11]"},{"why":"Provides the Wilson-loop characterization of inversion-symmetric topological bands used to read off the Chern number from winding.","marker":"[15]"},{"why":"Guarantees exponentially localized Wannier states for the trivial bands, grounding the real-space truncation.","marker":"[51]"}],"fun_headline_variants":["Interactions alone can forge Chern excitons from trivial bands","Making topological excitons without topological bands","Chern excitons from trivial bands via pure interactions","2D model shows interactions can induce topological excitons","Exciton alchemy: topological order from trivial bands"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The Chern number is computed after cutting the exciton's internal motion to the five shortest electron-hole separations, and the paper assumes this value survives when all larger separations are included.","fun_headline_variants_meta":{"raw":{"variants":["Interactions alone can forge Chern excitons from trivial bands","Making topological excitons without topological bands","Chern excitons from trivial bands via pure interactions","2D model shows interactions can induce topological excitons","Exciton alchemy: topological order from trivial bands"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000217,"raw_usage":{"total_tokens":1475,"prompt_tokens":1023,"completion_tokens":452,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":639,"completion_tokens_details":{"reasoning_tokens":378}},"tokens_in":639,"tokens_out":452,"duration_ms":4054,"temperature":1.0,"reasoning_tokens":378,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:27:37.950786+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Enlarge the relative-coordinate cutoff from $|\\Delta_x|,|\\Delta_y|\\le 10$ to systematically larger values and recompute the Wilson-loop winding of the lowest exciton band; if the winding ever departs from $C_{\\mathrm{exc}}=+1$, or the bulk gap closes against the exciton continuum before the infinite-basis limit, the topological claim fails.","supporting_citations":[],"review_version":2}