{"id":"814995a9-0484-4063-9934-ab466d19d6a6","arxiv_id":"2608.10747","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A reduced-flux composite-boson ansatz accounts for the full low-energy quasihole manifolds of the torus Hofstadter-Bose-Hubbard model at fillings nu<1/2.","lead":"Physicists propose an explicitly built composite-boson wave function that describes the low-energy manifolds of lattice bosons in a magnetic field at fillings below one half. They show the trial states reproduce the manifold dimension, energy spectra, fractional charge depletion, and braiding phase expected for fractional quantum Hall quasiholes.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"CB span is validated only inside lowest-band projection; full-Hamiltonian fidelity at U/t=2 is never computed, so the central identification may be an artifact of the projection.","rationale":"The reader's weakest assumption identifies exactly the load-bearing point: the equivalence between the projected CB span and the true low-energy manifold is presupposed, not demonstrated, for the U/t=2 model. My stress-test pass found no internal inconsistency in the counting or the theta-function construction; the orbit-resolved rank checks in Supplemental Table II are direct numerical validations of the cyclic-reduction prediction, and the subspace fidelities inside the projected model are strong. The missing piece is a direct check of the CB states against the full, unprojected HBH spectrum at the same parameters. This is an addressable omission rather than a demonstrated failure, so the conditional verdict remains appropriate. I therefore keep the reader's verdict unchanged while emphasizing that a full-ED fidelity test is the decisive next step.","tokens_in":28208,"tokens_out":11891,"duration_ms":126541,"concrete_test":"Perform full, unprojected exact diagonalization of the HBH Hamiltonian, Eq. (2), at U/t=2 for a small system such as (N,N_phi)=(2,8) on an 8x8 torus with phi=1/8. Identify the lowest D=20 states and confirm they form a gapped manifold. Evaluate the CB trial states of Eq. (6) directly in the full real-space Fock space (allowing double occupancy, no band projection), orthonormalize them, and compute the subspace fidelity F_sub = (1/D)||Q_CB^dag Q_exact||_F^2 and the worst principal angle. If F_sub < 0.99, the lowest-band projection is not innocent and the CB description of the physical U/t=2 model is unsupported. Repeat the same test with the hard-core restriction to separate the effect of U from the effect of projection.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that the composite-boson trial span equals the genuine low-energy manifold of the Hofstadter-Bose-Hubbard model at U/t=2. In 'Numerical Results,' both the 'exact' spectrum and the CB ansatz are projected onto the lowest Hofstadter band; the only unprojected calculation, Supplemental Material Sec. 5.2, uses the hard-core U->infinity Hamiltonian. Thus the CB wavefunctions are never compared with the full HBH low-energy manifold at the interaction strength actually used in the spectral comparisons. If lowest-band projection fails at phi=1/L1 and U/t=2 - e.g., because interband coupling or Landau-level mixing reshapes the low-energy subspace - then the reported subspace fidelities (F_sub > 0.9996 for the projected model) and the spectral agreement in Fig. 1 are artifacts of the projection, not evidence for the microscopic CB description. The VMC hard-core comparison does not control for this: it changes both the Hilbert-space restriction (removed) and the interaction (infinite U), so agreement with projected U/t=2 energies could be coincidental or reflect only a shared topological phase without the CB states being the actual physical wavefunctions. The failure of either the projection or the hard-core equivalence would break the rank, spectral, and topological identification that the central claim rests on.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a composite-boson (CB) trial wavefunction ansatz for the low-energy manifolds of the Hofstadter–Bose–Hubbard (HBH) model on a torus at fillings ν < 1/2. The ansatz attaches two vortices to each boson, reducing the flux to N_d = N_φ − 2N, and constructs trial states from occupation patterns of N bosons in N_d reduced-flux orbitals, multiplied by a squared Jastrow factor and a center-of-mass theta function. The authors show that organizing the occupation patterns into cyclic orbits and including the center-of-mass multiplicity reproduces the known degeneracy formula D = binom(N_d+N−1,N) N_φ/N_d. They validate the ansatz via orbit-resolved numerical ranks, subspace fidelities > 0.9996 for the lowest-band-projected Hamiltonian, spectral comparisons, many-body Chern numbers satisfying C_MB/D = ν, fractional density depletion Q = ν_eff, and an Aharonov–Bohm-free braiding phase consistent with 2ν_eff. VMC extends the rank and energetic analysis to larger systems in the hard-core limit.","tokens_in":28425,"tokens_out":11091,"duration_ms":108213,"significance":"If correct, the paper provides a rare microscopic wavefunction description for a broad family of lattice FQH manifolds, going beyond counting rules. The orbit-resolved rank tests are a strong, parameter-free check: each cyclic orbit's rank matches the prediction, including shortened orbits, and the predicted center-of-mass sectors are separately verified. The subspace fidelities above 0.9996 and consistent Chern numbers across many system sizes are convincing within the projected model. The use of VMC to test the same ansatz without explicit lowest-band projection in the hard-core limit is a valuable robustness check, and the disorder/asymmetric-pinning tests for the braiding phase are thoughtful. The main limitation is that the direct validation is performed within the lowest-band projection at U/t = 2, while the abstract and conclusion refer to the full HBH model without qualification.","major_comments":[{"comment":"The central claim that the CB span describes the low-energy manifolds of the HBH model at U/t = 2 is not directly tested against the full Hamiltonian. The exact spectra and the subspace fidelities in Table IV are computed for the lowest-band-projected Hamiltonian only, and the only unprojected calculation (Supplemental Sec. 5.2) uses the hard-core limit U → ∞, which changes both the Hilbert space and the interaction. Consequently, the agreement reported in Fig. 1 and in Table IV could in principle be an artifact of the lowest-band projection rather than evidence that the CB states describe the genuine low-energy manifold of Eq. (2). Because the abstract and conclusion state the framework applies to the HBH model without projecting, this gap is load-bearing. I request that, for at least the smaller systems (for example (N,N_φ) = (2,8) or (3,9)), the authors compute the overlap or subspace fidelity between the CB span and the low-energy manifold of the full HBH Hamiltonian at U/t = 2, or otherwise demonstrate quantitatively that the lowest-band projection is accurate at this interaction strength.","section":"Numerical Results; Supplemental Material Secs. 5.1 and 5.2"}],"minor_comments":[{"comment":"The ratio δ/Δ is used in the caption and text but is not defined in the caption; please define it there or refer to its definition in the main text.","section":"Fig. 1"},{"comment":"The sentence \"The number of nonzero eigenvalues of the trial-state overlap matrix confirms this physical rank\" does not state the numerical threshold for \"nonzero\"; the threshold appears later in Supplemental Sec. 2.4 (singular-value ratio > 10^-10) and should be mentioned here.","section":"Main text, above Eq. (5)"},{"comment":"The cyclic-orbit reduction is motivated by a continuous-translation identity that shifts all orbital labels by one, but the numerical trial states are generated by discrete one-site magnetic translations; the paper should state more explicitly that the orbit structure is a predicted reduction that is verified numerically rather than a mathematical proof for the discrete case.","section":"Supplemental Material Sec. 2.2"},{"comment":"The phrase \"N_loc added flux quanta are localized by repulsive potentials\" is potentially confusing; N_loc is implemented by changing L2 (and thus the total flux) and pinning the resulting quasiholes, so it would be clearer to say \"localized flux insertions\" instead of \"added flux quanta\" to avoid implying extra particles.","section":"Main text, 'Fractional density depletion' paragraph"},{"comment":"The column headers n_O and s_O are not defined in the table caption; they are defined in the surrounding text, but a brief definition in the caption would improve readability.","section":"Supplemental Material Table II"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically sound within its projected-model scope: the orbit-resolved rank tests, subspace fidelities, and Chern-number checks are convincing and well documented. The central issue is the mismatch between the stated claim (description of the full HBH model) and the validation (lowest-band-projected model with a separate hard-core VMC check). This is fixable by adding a full-Hamiltonian fidelity check for one or two small systems. The braiding phase result is honestly presented as the selection of an integer m; the robustness tests are appropriate. I do not see a circularity problem, because the rank prediction is tested independently of the degeneracy formula that motivated it."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a well-executed finite-size study that mostly delivers what it says, with one caveat worth keeping in front of you. The composite-boson identification with the HBH low-energy manifold is tested inside lowest-band projection and in hard-core VMC, but never in the full U/t=2 Hilbert space. The stress-test note has a real point, even though the paper is not hiding it.\n\nWhat's new: the unpinned reduced-flux orbital basis, the orbit-resolved rank analysis, and the AB-free braiding phase for generic sub-half-filled manifolds. The counting formula is a reorganization of earlier work (ref [32], which the authors cite), but the rank analysis is not circular: the overlap-matrix ranks are computed independently and match the orbit-by-orbit predictions, down to the shortened orbits. That is the strongest part of the paper. The subspace fidelities (F_sub > 0.9996) for the projected manifold, the Chern-number checks across N=2..5 with C_MB/D=nu, and the VMC extension to (5,11) and (6,14) with ranks 11 and 49 are all real evidence. The braiding diagnostics are also carefully monitored, with overlap, bandwidth-to-gap, disorder, and asymmetric-pinning checks.\n\nSoft spots, in order. First, the finite-U gap: the spectral and fidelity comparisons are all against the lowest-band-projected Hamiltonian at U/t=2; the only unprojected calculation is hard-core U to infinity. So the microscopic wavefunction description is not directly validated for the full HBH model at U/t=2. Lowest-band projection is standard and probably good at low flux, but the paper should either compute a direct unprojected comparison for a small system or soften the claim. Second, the braiding phase is determined for one large system, and the selection of the integer m is not independently pinned down there; the robustness checks are for N=2. Third, no code or data are shipped, and the Chern-number mesh is quoted at N_theta=21 without a convergence scan. These are minor-to-moderate and all addressable.\n\nBottom line: the central argument holds up for the projected model and for the hard-core model; the overextended step is the unqualified statement about the HBH model at finite U. This paper deserves serious refereeing. I would send it out, with a request for either the direct full-Hamiltonian check or a precise statement of what has been proven. For people working on lattice FQH or composite-particle trial states, it is worth citing and worth discussing.","headline":"Careful and mostly right, but the CB identification with the finite-U HBH manifold is only tested inside lowest-band projection; deserves review with a request for a direct full-Hamiltonian check.","tokens_in":29019,"tokens_out":3344,"would_cite":true,"duration_ms":32480,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81V70","82B20"],"pacs":["73.43.-f","73.43.Cd","67.85.-d"],"model":"deepseek-v4-flash","headline":"Attaching two vortices to each boson yields a trial wave-function basis that reproduces the low-energy quasihole manifolds of the Hofstadter–Bose–Hubbard model at every filling below one half.","keywords":["fractional quantum Hall effect","composite bosons","Hofstadter-Bose-Hubbard model","Laughlin quasiholes","torus theta functions","many-body Chern number","braiding phase","lattice bosons"],"falsifier":"A concrete check: pick a filling not listed, for example $(N,N_\\phi)=(4,11)$ or $(3,11)$, compute the exact lowest-band-projected low-energy spectrum and the overlap matrix of the composite-boson trial family, and see whether the numerical rank equals $\\binom{N_d+N-1}{N}N_\\phi/N_d$ and whether the worst principal angle with the exact manifold is as small as the $>0.9996$ fidelities reported here; any case where the rank is larger or the fidelity falls sharply would falsify the generic-filling claim.","tokens_in":27969,"feed_emoji":"🌀","tokens_out":10466,"duration_ms":97553,"temperature":0.7,"pith_summary":"This paper tries to show that the low-energy manifolds of the Hofstadter–Bose–Hubbard model at all fillings $\\nu<1/2$ can be described by a simple composite-boson wave-function basis. The idea is to attach two vortices to every boson, leaving $N_d=N_\\phi-2N$ effective orbitals, and to occupy those reduced-flux orbitals in every possible way, with magnetic translations trimming the naive occupation count to the known manifold dimension. The paper verifies that this basis spans the exact projected low-energy space with subspace fidelities above $0.9996$, matches its energy spectrum, and reproduces the many-body Chern number, fractional density depletion, and an Aharonov–Bohm-free braiding phase that identify these manifolds as lattice fractional quantum Hall states. If the ansatz is right, it converts a set of exclusion-rule counting formulas into explicit wave functions, making the quasihole structure of lattice fractional Hall systems directly manipulable.","feed_headline":"Two-vortex boson states reproduce lattice quantum Hall manifolds","feed_subtitle":"Attaching two vortices per boson matches spectra, Chern numbers, and quasihole braiding phases below half filling.","key_machinery":"The load-bearing object is the reduced-flux composite-boson ansatz: a many-body trial state built as a symmetrized product of $N$ single-particle torus $\\theta$-function orbitals at reduced flux $N_d=N_\\phi-2N$, multiplied by the square of the odd Jacobi $\\theta$ function $\\vartheta_{1/2,1/2}((z_i-z_j)/L_1|\\tau)$ and by a center-of-mass $\\theta$ factor $F_{\\mathrm{CM}}(Z)=\\vartheta_{a,0}(2Z/L_1|2\\tau)$. The counting mechanism is the organization of occupation patterns of $N$ bosons in $N_d$ orbitals into cyclic orbits under a common shift of all orbital labels; each orbit of size $s$ contributes $s/n$ internal directions with $n=N_d/g$ and $g=\\gcd(N,N_d)$, and the $q_{\\mathrm{COM}}=N_\\phi/g$ center-of-mass translations multiply the count, yielding $D=\\binom{N_d+N-1}{N}N_\\phi/N_d$. Overlap-matrix ranks verify this prediction orbit by orbit, so the machinery converts bosonic occupations of reduced-flux orbitals into a basis for the physical low-energy manifold.","core_discovery":"The central claim is that every low-energy manifold of the Hofstadter–Bose–Hubbard model at fillings $\\nu<1/2$ has an explicit microscopic wave-function description in terms of composite bosons. Starting from $N_\\phi$ flux quanta and $N$ bosons, attaching two vortices per boson leaves $N_d=N_\\phi-2N$ reduced-flux orbitals; the trial states are symmetrized products of $\\theta$-function orbitals at this reduced flux, multiplied by the squared odd Jacobi $\\theta$ function $\\vartheta^2_{1/2,1/2}((z_i-z_j)/L_1|\\tau)$ and a center-of-mass $\\theta$ factor, with the whole reference state complex-conjugated to match the physical Hofstadter chirality. The naive occupation count $\\binom{N_d+N-1}{N}$ is overcomplete, but organizing occupation patterns into cyclic orbits under a common shift of orbital labels and including the $q_{\\mathrm{COM}}=N_\\phi/\\gcd(N,N_d)$ center-of-mass sectors predicts the physical rank $D=\\binom{N_d+N-1}{N}N_\\phi/N_d$. Numerical overlap matrices confirm this rank orbit by orbit for the systems $(N,N_\\phi)=(2,8),(2,10),(3,9),(3,10)$, subspace fidelities with the exact projected low-energy manifold exceed $0.9996$, and variational Monte Carlo finds the predicted ranks $D=11$ and $49$ for $(5,11)$ and $(6,14)$. The same manifolds carry Chern number $C_{\\mathrm{MB}}/D=\\nu$, pinned added-flux excitations deplete the density by $\\nu_{\\mathrm{eff}}$, and an Aharonov–Bohm-free braid gives phase $2\\nu_{\\mathrm{eff}}$.","pith_inferences":["Inference: if the reduced-flux occupation organization is the correct internal label for these manifolds, the same occupations should predict finer structure, such as the decomposition of the entanglement spectrum or per-sector Berry curvature, rather than only the total dimension; this is testable in the reported systems.","Inference: because the braiding phase and depletion track $\\nu_{\\mathrm{eff}}$ rather than a fixed denominator, the finite-size manifolds behave as Laughlin-quasihole multiplets at an effective filling; explicit adiabatic continuity between these lattice states and the continuum quasihole sector would be a natural check not performed in the paper.","Inference: the same cyclic-orbit counting should apply to fermionic Hofstadter models with an odd number of attached vortices, where a composite-fermion analogue would predict the rank of Jain-type quasihole manifolds; the paper treats only two-vortex bosons.","Inference: pinning and braiding protocols could be designed using the composite-boson occupations as a guide, for example choosing pin configurations that isolate a single reduced-flux occupation sector, which may simplify future experimental or numerical probes of these lattice states."],"forward_implications":["The known quasihole counting formula for the sub-half-filled Hofstadter–Bose–Hubbard model is realized by explicit wave functions: each low-energy manifold is spanned by the composite-boson trial family, not just counted by it.","For the systems checked, the composite-boson span reproduces the lowest-band-projected spectrum and the exact low-energy subspace, so the ansatz can serve as a variational tool for lattice fractional quantum Hall states.","The unpinned manifolds carry total many-body Chern number $C_{\\mathrm{MB}}=\\binom{N_d+N-1}{N_d}$, so the averaged Chern number per state is exactly the filling $\\nu=N/N_\\phi$, confirming their topological character.","A localized added-flux excitation inside these manifolds depletes the density by $Q=\\nu_{\\mathrm{eff}}$ and acquires a braiding phase $2\\nu_{\\mathrm{eff}}$ in an Aharonov–Bohm-free loop, meaning the quasihole response tracks the effective filling.","Variational Monte Carlo evaluation without explicit lowest-band projection finds the predicted ranks and narrow manifolds for $(N,N_\\phi)=(5,11)$ and $(6,14)$, indicating the ansatz extends beyond the smallest systems."],"supporting_citations":[{"why":"Supplies the degeneracy formula and fractional-depletion diagnostics that the trial basis must reproduce.","marker":"[32]"},{"why":"Supplies the torus theta-function boundary conditions and Laughlin–Jastrow wave functions used to build the trial states.","marker":"[44]"},{"why":"Supplies the torus many-body wave functions and center-of-mass factor construction extended here.","marker":"[45]"},{"why":"Gives the generalized Pauli-principle quasihole counting that the ansatz rank matches.","marker":"[20]"},{"why":"Gives the thin-torus counting argument whose wave-function counterpart the ansatz provides.","marker":"[47]"},{"why":"Supplies composite-particle projection and ansatz-subspace diagonalization methods used to test the trial basis.","marker":"[12]"},{"why":"Documents the linear dependence of composite-particle trial spaces that motivates the rank reduction by cyclic orbits.","marker":"[67]"},{"why":"Defines the Laughlin quasihole braiding phase $2/m$ that the measured $2\\nu_{\\mathrm{eff}}$ is compared with.","marker":"[7]"},{"why":"Supplies the twist-angle Berry-phase method used for the many-body Chern number.","marker":"[30]"},{"why":"Adapts the many-body Chern-number method to lattice systems with projected bands.","marker":"[31]"}],"fun_headline_variants":["Composite-boson ansatz for lattice FQH at any sub-half filling","Two-vortex bosons reproduce spectra and Chern numbers on a lattice","Lattice bosons: composite-boson wavefunctions match FQH manifolds","Composite-boson ansatz explains lattice FQH below half filling","Two-vortex attachment captures lattice boson FQH manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that projecting the Hofstadter–Bose–Hubbard model onto its lowest band at $U/t=2$, or replacing it by the hard-core real-space limit in the Monte Carlo comparison, leaves the low-energy manifold essentially intact, so that the continuum theta-function trial states evaluated on lattice sites still span the physically relevant space.","fun_headline_variants_meta":{"raw":{"variants":["Composite-boson ansatz for lattice FQH at any sub-half filling","Two-vortex bosons reproduce spectra and Chern numbers on a lattice","Lattice bosons: composite-boson wavefunctions match FQH manifolds","Composite-boson ansatz explains lattice FQH below half filling","Two-vortex attachment captures lattice boson FQH manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000822,"raw_usage":{"total_tokens":3723,"prompt_tokens":1198,"completion_tokens":2525,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":814,"completion_tokens_details":{"reasoning_tokens":2426}},"tokens_in":814,"tokens_out":2525,"duration_ms":18120,"temperature":1.0,"reasoning_tokens":2426,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T18:09:05.399900+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: pick a filling not listed, for example $(N,N_\\phi)=(4,11)$ or $(3,11)$, compute the exact lowest-band-projected low-energy spectrum and the overlap matrix of the composite-boson trial family, and see whether the numerical rank equals $\\binom{N_d+N-1}{N}N_\\phi/N_d$ and whether the worst principal angle with the exact manifold is as small as the $>0.9996$ fidelities reported here; any case where the rank is larger or the fidelity falls sharply would falsify the generic-filling claim.","supporting_citations":[{"cited_title":"Greiter, V","cited_arxiv_id":null,"evidence_quote":"Supplies the torus many-body wave functions and center-of-mass factor construction extended here."},{"cited_title":"Seidel and K","cited_arxiv_id":null,"evidence_quote":"Gives the thin-torus counting argument whose wave-function counterpart the ansatz provides."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Adapts the many-body Chern-number method to lattice systems with projected bands."}],"review_version":1}