{"id":"217f50fa-cfa5-4b37-9ebf-196e8af3ac17","arxiv_id":"2411.19287","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Interactions reduce the free-fermion crystalline SPT classification for U(1) with reflection, rotation, or inversion symmetries to finite quotients: Z4 in 1D, Zn-based groups in 2D, and Z8 times Z2 in 3D.","lead":"The authors use the Atiyah-Hirzebruch spectral sequence to relate free-fermion and interacting classifications of crystalline symmetry-protected topological phases, and they compute the interaction effects in several explicit examples. A general reader might care because it offers a systematic recipe for deciding which topological phases remain robust when electron interactions are switched on.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (17), ∂∘κ=κ∘∂, is asserted without proof and is load-bearing: it licenses κ^r as a map of AHSS pages. Directly checking that the quantum-number replacements commute with d1 in the 1D, 2D, and 3D examples would settle whether Tables 1–3 are justified.","rationale":"I agree with the reader that Eq. (17) is the weakest point. The paper's whole mechanism is to quotient free-fermion E∞ classes by interacting data; the only bridge between the two spectral sequences is κ, and the only justification for its compatibility with differentials is a one-line ‘physically reasonable’ statement. The examples may well be correct, and the Appendix A lattice models give independent support for the 1D reduction; the explicit check I propose would turn the conditional into an acceptance-style verification. I recommend no change to the reader's CONDITIONAL verdict, because the concern does not point to a known contradiction but to a missing proof of a central hypothesis.","tokens_in":23447,"tokens_out":11598,"duration_ms":108256,"concrete_test":"Test the E1-page chain-map condition in each example. For 1D reflection: free d1 sends the 1-cell generator to (1,1)_f; applying κ via Eq. (41) gives (2,1)_I, which is exactly the interacting d1 in Eq. (34), so the square commutes on the generator. Repeat for C_n rotation using Eq. (49), the assignment (70), and the interacting d1 of Eq. (58); and for 3D inversion compare κ(d1^F) with d1^I on the 3-cell generator. In addition, verify the non-split 3D relation Eq. (93) by an explicit lattice/interacting construction of two Chern-insulator layers and checking that its many-body ground state has the same (N_2-cell, N_BIQH) quantum numbers as the claimed monopole-charge state. If any of these checks fails, Eq. (17) or the underlying quantum-number identification must be amended; if all pass, the reductions in Tables 1–3 are supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.4 states that κ: K_0^G(X,Y)→h_0^G(X,Y) is compatible with the boundary homomorphism and writes ∂∘κ=κ∘∂. From this it immediately concludes that κ induces maps κ^r on every AHSS page and that all differentials commute with κ. That inference is the crux of the paper: Tables 1–3 are obtained by applying these page maps and then quotienting free classes by interacting d1 images and extension relations. Naturality of the global boundary map for one pair (X,Y) does not, by itself, prove naturality of the filtration boundary maps that define the AHSS differentials; one needs κ to be a natural transformation of the underlying homology theories, or an explicit proof for the filtration pairs (X_p, X_{p-1}). No such proof is given; the paper calls (17) ‘physically reasonable.’ If (17) failed in any example, the per-cell maps could not be assembled, and the reductions [N_R]→[N_R mod 4], Table 2’s mod-n groups, and [N_2-cell]→[(N_2-cell mod 8,0)] would not follow from the stated framework. This is a correctness risk, not a stylistic gap.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes an Atiyah-Hirzebruch spectral sequence (AHSS) framework for connecting free-fermion and interacting crystalline symmetry-protected topological (SPT) phases. The central idea is a homomorphism κ from K-homology to generalized homology that is claimed to commute with the bulk-boundary boundary map and hence with all AHSS differentials, inducing maps between spectral sequence pages. Working mainly with E1-page quantum numbers, the authors derive explicit reduction rules: 1d U(1)×Z_2^R free-fermion classes [N_R] reduce to [N_R mod 4]; 2d U(1)×C_n rotation classes reduce to mod-n or mod-2n/2 groups plus a Chern number factor; and 3d U(1)×inversion classes [N_{2-cell}] reduce to [(N_{2-cell} mod 8, 0)]. The paper includes a 1d lattice-model appendix supporting the 1d reduction and emphasizes that the approach handles both split and non-split short exact sequences.","tokens_in":23720,"tokens_out":4109,"duration_ms":37457,"significance":"If the framework is correct, it offers a systematic and computationally tractable route from free-fermion classifications to interacting crystalline SPT classifications, a topic that has so far been addressed only in isolated examples. The paper's explicit predictions (mod-4 in 1d, mod-n and mod-2n in 2d, mod-8 in 3d) are concrete and falsifiable, and the appendix gives a family of lattice models realizing the 1d reduction. The AHSS perspective, using the same spectral sequence for both K-homology and interacting generalized homology, is a valuable organizing principle. The main reservations concern not the results' plausibility but the proof infrastructure: the compatibility condition that makes the whole page-by-page comparison possible is asserted rather than proved, and one stated structural assumption is contradicted by the paper's own 2d and 3d examples.","major_comments":[{"comment":"The condition ∂∘κ=κ∘∂ is asserted as 'physically reasonable' but is never proved. This condition is load-bearing: it is exactly what licenses the induced maps κ^r between all AHSS pages and justifies the commutative diagram (20) and every reduction in Tables 1–3. Naturality of the boundary map for one pair (X,Y) does not automatically give naturality of the filtration boundary maps defining the spectral-sequence differentials; one needs a proof that κ is a natural transformation of the relevant homology theories, or at least an explicit verification for the filtration pairs (X_p, X_{p-1}). Without such an argument, the per-cell quantum-number replacements cannot be assembled into a well-defined map on equivalence classes. A direct check of ∂∘κ=κ∘∂ for the 1d, 2d, and 3d examples would settle the issue and would be within the paper's scope.","section":"Sec. 2.4, Eq. (17)"},{"comment":"The paper states that for p=2 it 'consider[s] cases where there is no reduction, i.e., κ^1_{2,-2} is an isomorphism,' but in the 2d example the free E^1_{2,-2} is Z (Chern number) while the interacting E^1_{2,-2} is Z_2 (Chern plus bosonic integer quantum Hall states), and in the 3d example the same entry is Z versus Z_2. These are not isomorphic, and the paper itself later sends [N_CI] to [(N_CI,0)] (Eq. (68)) and [N_{2-cell}] to [(N_{2-cell} mod 8, 0)] (Eq. (98)), which is not an isomorphism. This inconsistency undermines the stated structural assumption and the derivation of the commutative diagram (20) in the very cases where the examples are worked out. The assumption should be replaced with a correct statement of what κ^1_{2,-2} actually is in these examples, with a computation.","section":"Sec. 2.4 and Sec. 4.3/5.3"},{"comment":"The non-split extension in 3d rests on the deformation equivalence (N_CI=2)_f ≅ (n_+=1,n_-=0)_f, and the interacting version (N_CI=2,N_BIQH)_I ≅ (N_C=1,I_C=0)_I ⊕ (N_BIQH)_I, is asserted with 'can be verified' and a citation to [53,55]. This equivalence is load-bearing: it is what turns the short exact sequence (85)/(92) from a direct sum into Z and Z_8×Z_2, respectively. Since the paper presents the computation as its own framework rather than as a direct quotation, a proof or a precise theorem statement with the exact statement being imported from [53,55] should be supplied. Without it, the non-split example and the final mod-8 reduction rule do not follow from the arguments given.","section":"Sec. 5.1, Eq. (93)"}],"minor_comments":[{"comment":"The tables labeled 'E1' in Eqs. (59), (60), and (61) are E2 pages, not E1 pages, based on the surrounding text ('Using these results to compute the E2-page'); the labels should be corrected.","section":"Sec. 4.2, Eqs. (59)–(61)"},{"comment":"The phrase 'pertaining' is misspelled as 'pertainning' in Sec. 4.3, and similar small typographical errors appear elsewhere; the manuscript would benefit from a careful proofreading pass.","section":"Sec. 3.3 and Sec. 4.3"},{"comment":"The notation κ^2_{0,0} and κ^2_{2,-2} is introduced without explicitly defining the superscript as the page index; adding one sentence explaining the notation would improve readability.","section":"Sec. 2.4, Eq. (20)"},{"comment":"The E1-page table for the free-fermion 2d case shows entries 'Z' at (p=1,q=0) and (p=2,q=0), but the surrounding discussion only explicitly defines the differential d^1_{1,0}; the reader must infer d^1_{2,0}=0. A sentence stating this explicitly would help.","section":"Sec. 4.1, Eq. (46)"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on three prior works by the same research group ([53,55,59]) for the local E1-page data, the non-split deformation equivalence, and the decomposable-system quantum numbers. This is not by itself a flaw, but it means the independence of the present verification is limited; the editor may wish to ensure that the cited results are indeed published and that the imported statements are correctly represented. The main revision requested—proving or precisely citing the naturality condition Eq. (17) and correcting the κ^1_{2,-2} assumption—belongs to the authors' scope and, if carried out, could make the framework sound."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper gives a concrete way to connect free-fermion and interacting crystalline SPT classifications through the AHSS, and the main results—1D reflection reducing Z to Z4, 2D Cn rotations reducing to mod-n groups, 3D inversion reducing to Z8—are explicit and, as far as I can see, correct. What is genuinely new is the κ map between the quantum numbers on each side and the assembled reduction tables. The framework itself is borrowed from prior AHSS work, but this paper makes the interaction-reduction computable for split and non-split extensions with worked examples and a lattice-model check in Appendix A. That is real value.\n\nThe soft spot is the one the paper itself flags: Eq. (17), ∂∘κ=κ∘∂, is asserted as 'physically reasonable' and then used to define κ^r on every AHSS page. Naturality of the global boundary map does not automatically give naturality of the filtration boundary maps, so the page maps are not fully justified. If that compatibility fails in some example, the per-cell maps cannot be assembled and Tables 1–3 don't follow. This is a correctness risk, not a stylistic gap. A referee should ask for a proof, or at minimum a direct check that the quantum-number replacement commutes with d1 in each of the three examples.\n\nThere is also an internal inconsistency in Sec 5.1: the text says there are no higher differentials and then invokes d2_2,-2 to argue the non-split extension. That needs clarification. And the E1 pages and deformation equivalence are imported from [53,55], which are published and by the same group; acceptable, but it means the derivation is not self-contained.\n\nOverall, the central argument holds up for the examples if Eq. (17) is granted. This deserves a serious referee. My recommendation: send it out, and have the referee focus on (17), the d2 usage, and the misprint in the 3D differential.","headline":"A useful, plausible spectral-sequence map between free and interacting crystalline SPT classifications; the main tables are likely right, but the load-bearing compatibility assumption (Eq. 17) is asserted, not proven.","tokens_in":24278,"tokens_out":2528,"would_cite":true,"duration_ms":21600,"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":"The paper establishes that interactions quotient free-fermion crystalline SPT classifications down to finite abelian groups: $\\mathbb{Z}\\to\\mathbb{Z}_4$ in 1D reflection, mod-$n$ reductions under $C_n$ rotations, and…","keywords":["crystalline symmetry-protected topological phases","free-fermion classification","interaction effects","Atiyah-Hirzebruch spectral sequence","K-homology","generalized homology","point-group symmetry","bulk-boundary correspondence"],"falsifier":"Take one of the decomposable lattice models $H_{p,n}$ from Appendix A, add strong local interactions, and determine the many-body class exactly or numerically: the paper predicts $[n-p \\bmod 4]$ for every $(p,n)$. A single mismatch with the interacting classification would show that $\\partial\\circ\\kappa=\\kappa\\circ\\partial$ fails and the reductions do not follow. Alternatively, look for a free-fermion phase whose boundary state, mapped to the interacting side, is not the image of the interacting boundary map.","tokens_in":23245,"feed_emoji":"⚛️","tokens_out":12842,"duration_ms":97996,"temperature":0.7,"pith_summary":"The paper's aim is to turn a hard question — what happens to free-fermion topological classifications when interactions are switched on — into spectral-sequence bookkeeping. Its central claim is that the natural map from free-fermion phases to interacting phases commutes with every step of the Atiyah-Hirzebruch spectral sequence, so interactions simply quotient the free-fermion integer invariants by the image of certain cell-pumping maps. For 1D reflection-symmetric systems this gives $\\mathbb{Z} \\to \\mathbb{Z}_4$; for 2D $C_n$-symmetric systems it gives the finite reductions collected in Table 2; for 3D inversion-symmetric systems it gives $\\mathbb{Z} \\to \\mathbb{Z}_8 \\times \\mathbb{Z}_2$. A reader should care because the reductions are explicit, computable, and checkable on small lattice models: interaction effects stop being a case-by-case mystery and become a quotient.","feed_headline":"Interactions shrink free-fermion phases to mod-4 and mod-8 classes","feed_subtitle":"A spectral-sequence comparison turns interaction effects into explicit quotients in 1D, 2D, and 3D.","key_machinery":"The engine is the Atiyah-Hirzebruch spectral sequence (AHSS), a bookkeeping device that decomposes the real-space manifold into $G$-symmetric cells and computes the full homology class from local SPT data on each cell by repeatedly taking kernels and images of differentials. The paper's specific machinery is the induced homomorphism $\\kappa^r$ between the free-fermion pages $E^{\\mathrm{free},r}$ and the interacting pages $E^r$; the first differential describes pumping SPT states from a cell onto lower-dimensional cells, and the quantum numbers $(n_+,n_-)$, $(N_C,R_C)$, and $(N_{2\\text{-cell}},N_{\\rm BIQH})$ encode which cell states survive the quotient. The commuting diagram (20) then assembles the per-cell comparisons into a map on the final classifications.","core_discovery":"The paper's central claim is that the free-fermion classification of crystalline SPT phases, $K_0^G(X,Y)$ with $G$ a point group, is related to the interacting classification $h_0^G(X,Y)$ by a homomorphism $\\kappa\\colon K_0^G(X,Y)\\to h_0^G(X,Y)$. The key statement is the compatibility condition $\\partial \\circ \\kappa = \\kappa \\circ \\partial$ (Eq. (17)), which asserts that mapping a free-fermion phase to its many-body incarnation does not interfere with bulk-boundary correspondence. Granting this, $\\kappa$ induces maps on every page of the two AHSS computations, and the spectral-sequence differentials commute with $\\kappa$. The reductions follow: $[N_R] \\mapsto [N_R \\bmod 4]$ in 1D; $[N_1]\\oplus[N_{\\rm CI}] \\mapsto [-N_1 \\bmod 4]\\oplus[(N_{\\rm CI},0)]$ for $C_2$, with analogous mod-$n$ formulas for $C_n$; and $[N_{2\\text{-cell}}] \\mapsto [(N_{2\\text{-cell}} \\bmod 8, 0)]$ for inversion in 3D. Thus interactions act as a computable quotient of the free-fermion equivalence classes.","pith_inferences":["Editorial inference: the same quantum-number comparison should yield concrete mod-2 and mod-4 predictions for other point groups, such as mirror Chern insulators with $U(1)$ and $C_s$ symmetry, whenever both $E^1$ pages are computed.","Editorial inference: because the reduction is built from cell-local data, it is natural to expect extensions to magnetic space groups and to higher-order topology; proving the commuting condition (17) in those settings would be the essential first step.","Editorial inference: the decomposable lattice models $H_{p,n}$ in Appendix A can be promoted to fully interacting numerical tests; a failure of the predicted $[n-p \\bmod 4]$ class under strong interactions would pinpoint where the commuting-differential assumption breaks."],"forward_implications":["In 1D systems with $U(1)$ and reflection symmetry, a free-fermion phase labeled by $N_R\\in\\mathbb{Z}$ becomes $[N_R \\bmod 4]$ once interactions are included (Eq. (43)).","In 2D systems with $U(1)$ and $C_n$ rotation symmetry, the free-fermion classes $[(N_1,\\dots,N_{n-1})]\\oplus[N_{\\rm CI}]$ map to the finite groups in Table 2, with the Chern number $N_{\\rm CI}$ unchanged but the bosonic integer quantum Hall coordinate forced to zero.","In 3D systems with $U(1)$ and inversion symmetry, the free-fermion $\\mathbb{Z}$ class $[N_{2\\text{-cell}}]$ reduces to $[(N_{2\\text{-cell}} \\bmod 8, 0)]$ inside the interacting $\\mathbb{Z}_8\\times\\mathbb{Z}_2$ classification (Eq. (98)).","The method handles both split and non-split group-extension short exact sequences, covering cases where stacking two trivial-looking pieces produces a nontrivial phase through an extension rather than a direct sum.","Because the reductions are phrased in terms of cell-local quantum numbers, the same comparison can be repeated for any symmetry whose free and interacting $E^1$ pages are known."],"supporting_citations":[{"why":"Supplies the interacting-side framework: generalized homology and the AHSS computation of crystalline SPT classifications, including the E1-page entries used here.","marker":"[53]"},{"why":"Supplies the free-fermion side: the AHSS in band topology used to compute the free-fermion pages.","marker":"[54]"},{"why":"Supplies the real-space AHSS calculations of free-fermion point-group classifications that give the E1 entries for reflection, rotation, and inversion examples.","marker":"[55]"},{"why":"Establishes the canonical interaction reduction from $\\mathbb{Z}$ to $\\mathbb{Z}_8$ for stacked Kitaev chains, the phenomenon this paper generalizes to crystalline symmetries.","marker":"[26]"},{"why":"Introduces the construction of crystalline topological phases from lower-dimensional states, the physical basis for the cell-decomposition and extensive-trivialization picture.","marker":"[47]"},{"why":"Provides the prior interacting classifications for 2D rotation-symmetric systems, the comparison case for the $C_n$ reductions.","marker":"[45]"},{"why":"Introduces decomposable systems and the quantum numbers used in Appendix A to make the free-to-interacting reduction concrete on lattice models.","marker":"[59]"},{"why":"Supplies the bosonic integer quantum Hall states that appear in the interacting E1 entries and are set to the zero coordinate in the reductions.","marker":"[58]"}],"fun_headline_variants":["Spectral sequence ties free-fermion to interacting SPT phases","Interactions quotient free-fermion SPT classes via spectral sequence","Interactions collapse free-fermion SPT to mod-4, mod-8","AHSS reveals interaction quotients in crystalline SPT phases","Free-fermion SPT phases reduce to mod-4 and mod-8 under interactions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the unproven compatibility condition $\\partial \\circ \\kappa = \\kappa \\circ \\partial$ stated in Sec. 2.4 (Eq. (17)): the map that turns a free-fermion phase into its interacting counterpart must respect the way a bulk is related to its boundary; the paper calls this physically reasonable but gives no argument, and if it fails the per-cell maps cannot be assembled into a well-defined map on equivalence classes.","fun_headline_variants_meta":{"raw":{"variants":["Spectral sequence ties free-fermion to interacting SPT phases","Interactions quotient free-fermion SPT classes via spectral sequence","Interactions collapse free-fermion SPT to mod-4, mod-8","AHSS reveals interaction quotients in crystalline SPT phases","Free-fermion SPT phases reduce to mod-4 and mod-8 under interactions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000726,"raw_usage":{"total_tokens":3261,"prompt_tokens":961,"completion_tokens":2300,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":2202}},"tokens_in":577,"tokens_out":2300,"duration_ms":14636,"temperature":1.0,"reasoning_tokens":2202,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:19:59.793464+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take one of the decomposable lattice models $H_{p,n}$ from Appendix A, add strong local interactions, and determine the many-body class exactly or numerically: the paper predicts $[n-p \\bmod 4]$ for every $(p,n)$. A single mismatch with the interacting classification would show that $\\partial\\circ\\kappa=\\kappa\\circ\\partial$ fails and the reductions do not follow. Alternatively, look for a free-fermion phase whose boundary state, mapped to the interacting side, is not the image of the interacting boundary map.","supporting_citations":[{"cited_title":"Generalized homology and atiyah–hirzebruch spectral sequence in crystalline symmetry protected topological phenomena","cited_arxiv_id":null,"evidence_quote":"Supplies the interacting-side framework: generalized homology and the AHSS computation of crystalline SPT classifications, including the E1-page entries used here."},{"cited_title":"Atiyah-hirzebruch spectral sequence in band topology: General formalism and topological invariants for 230 space groups","cited_arxiv_id":null,"evidence_quote":"Supplies the free-fermion side: the AHSS in band topology used to compute the free-fermion pages."},{"cited_title":"Topological classification under nonmag- netic and magnetic point group symmetry: Application of real-space atiyah-hirzebruch spectral sequence to higher-order topology","cited_arxiv_id":null,"evidence_quote":"Supplies the real-space AHSS calculations of free-fermion point-group classifications that give the E1 entries for reflection, rotation, and inversion examples."},{"cited_title":"Effects of interactions on the topological classification of free fermion systems","cited_arxiv_id":null,"evidence_quote":"Establishes the canonical interaction reduction from $\\mathbb{Z}$ to $\\mathbb{Z}_8$ for stacked Kitaev chains, the phenomenon this paper generalizes to crystalline symmetries."},{"cited_title":"Topological phases protected by point group symmetry","cited_arxiv_id":null,"evidence_quote":"Introduces the construction of crystalline topological phases from lower-dimensional states, the physical basis for the cell-decomposition and extensive-trivialization picture."},{"cited_title":"Characterization and classi- fication of interacting (2 + 1)-dimensional topological crystalline insulators with orientation- preserving wallpaper groups","cited_arxiv_id":null,"evidence_quote":"Provides the prior interacting classifications for 2D rotation-symmetric systems, the comparison case for the $C_n$ reductions."},{"cited_title":"Crystalline-equivalent topological phases of many-body fermionic systems in 1+1 dimensions","cited_arxiv_id":null,"evidence_quote":"Introduces decomposable systems and the quantum numbers used in Appendix A to make the free-to-interacting reduction concrete on lattice models."},{"cited_title":"Theory and classification of interacting integer topo- logical phases in two dimensions: A chern-simons approach","cited_arxiv_id":null,"evidence_quote":"Supplies the bosonic integer quantum Hall states that appear in the interacting E1 entries and are set to the zero coordinate in the reductions."}],"review_version":1}