{"id":"e208383a-67c8-4ef5-93fa-462262a0122c","arxiv_id":"2608.12455","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Exact lattice fermionic symmetries in (1+1)D and (2+1)D are captured by a hierarchy of cohomological anomaly indices that does not fully match the continuum QFT anomaly classification; a constructed Z_4^F symmetry blocks SRE states despite a trivial continuum anomaly.","lead":"The paper classifies fermionic 't Hooft anomalies that finite internal symmetries can have when they act exactly on lattice systems in one and two spatial dimensions. It shows some anomalies allowed in continuum quantum field theory cannot be realized on the lattice, and gives a concrete two-dimensional example where a lattice symmetry forbids a trivial gapped state even though its continuum anomaly is trivial.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The (2+1)D Majorana-layer index and the Z_4^F example depend on an unpublished theorem that all 2D fermionic QCAs are circuits; the explicit bT operator in Eq. (80) is not manifestly a circuit, so n2 may be ill-defined.","rationale":"The paper's strongest claim is the Sec. 7 mismatch between lattice and continuum anomalies. The logic is clear: an explicit Z_4^F operator with U_g^2=(-1)^F is constructed, its truncation gives a boundary Majorana translation with index sqrt(2), so n2=omega_2, and the Sec. 6 obstruction argument rules out symmetric SRE states. The weakest point is the well-definedness of n2, which is the foundation of the whole (2+1)D construction. The reader identified this correctly. I agree: the unnamed 'To appear' theorem is the load-bearing step, and the example's own bT operator is a concrete place where the theorem's applicability is doubtful. I do not see a stronger internal inconsistency; the continuum-triviality claim for Z_4^F is standard, the stacking and gauge transformations seem consistent, and the 1D results have their own support. Hence the concern does not overturn the verdict but keeps it conditional: the central conclusion is compelling if the unpublished theorem is valid in the needed generality, and otherwise the example's index is unproven. The recommended verdict remains CONDITIONAL, so I mark UNCHANGED relative to the reader.","tokens_in":21472,"tokens_out":20832,"duration_ms":191183,"concrete_test":"When the manuscript or an independent source supplies [56], check whether its theorem covers arbitrary locality-preserving unitaries (including controlled, non-translation-invariant ones such as bT) or only translation-invariant Clifford FQCAs; if the latter, the use in Sec. 5.1 is invalid. Independently, compute the minimal circuit depth needed to implement U_g of Eq. (80) on a finite honeycomb patch with a single square domain-wall loop of side length L: if the depth grows with L (e.g., by exact numerical decomposition into local gates for L=2,4,8), then U_g is not FDQC and the truncation that defines n2 in Sec. 7.1 is not justified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Sec. 5.1 defines the Majorana-layer index n2 by truncating the symmetry operator U(g) within a disk and reading the index of the resulting boundary operator Omega_{dA}. This truncation procedure is only meaningful for finite-depth circuits (FDQC), and the paper asserts without proof, citing the unpublished [56], that every fermionic QCA in 2D is trivial up to circuit multiplications and hence that U(g) is FDQC. The central Z_4^F example in Sec. 7.1 does not visibly satisfy this: U_g in Eq. (80) includes bT[{sigma^z_p}], a QCA that translates Majoranas along domain-wall loops of arbitrary length. For a single loop of perimeter L, the restriction of U_g to a disk appears to require a circuit of depth growing with L, so U_g is not manifestly FDQC. If [56] does not cover this case, the truncation U_A(g), the boundary operator Omega_{dA}, and the index n2=omega_2 are not defined by the paper's own prescription, and the proof in Sec. 6.1.1 that n2 obstructs SRE does not apply. The 1D assertion that finite internal symmetries are generated by FDQC is likewise assumed. The mismatch conclusion therefore rests on an unverifiable theorem.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a lattice characterization of fermionic 't Hooft anomalies for finite internal symmetries in (1+1)D and (2+1)D, formulated in terms of obstructions to symmetric short-range-entangled (SRE) states. In (1+1)D the authors define an Else--Nayak-type pair of anomaly indices (n2, ν3), derive their gauge transformations and stacking rules, and prove for G_f = G_b × Z_2^F that trivial indices imply the symmetry is onsiteable after adding ancillas. They conclude that exact lattice symmetries realize only a subgroup of the continuum anomaly classification, for example only the even Z_4 subgroup of the Z_8 classification for G_b = Z_2. In (2+1)D they define three anomaly layers (n2, n3, ν4), prove that a nontrivial value of any layer obstructs a symmetric SRE state, and prove a stronger obstruction to symmetric invertible states when the extension class ω2 is trivial. The central example is a Z_4^F lattice symmetry in (2+1)D carrying n2 = ω2, which they argue forbids an SRE state even though its continuum 't Hooft anomaly is trivial, thereby demonstrating a mismatch between lattice and continuum fermionic anomalies.","tokens_in":21741,"tokens_out":26831,"duration_ms":250228,"significance":"If the main claims hold, this is a valuable and timely contribution. It gives a concrete lattice framework for extracting fermionic anomaly data directly from symmetry operators, and it provides a nontrivial example in which an exact lattice symmetry obstructs an SRE state despite having a trivial continuum anomaly. The 1+1D completeness result (trivial lattice indices imply onsiteability) is also useful and carefully argued. The paper is generally careful: cocycle identities, gauge transformations, stacking rules, and obstruction proofs are written out in detail, and the authors are explicit about which statements are conjectural. The main weakness is that the most striking 2+1D claim rests on an unpublished and currently unverifiable theorem about 2D fermionic quantum cellular automata, and that theorem is load-bearing for the construction of the Majorana-layer index and for the Z_4^F example.","major_comments":[{"comment":"The definition of the Majorana-layer index n2 and the computation for the Z_4^F example both require the symmetry operator U(g) to be a finite-depth quantum circuit, so that the disk truncation U_A(g) in Eq. (41) and the boundary operator Ω_{∂A} are well defined. The paper justifies this with the statement 'the fermionic QCA in 2D space is trivial up to circuit multiplications [56]', where [56] is an unpublished manuscript ('J. Haah, To appear'). This is not merely a citation problem: the explicit Z_4^F generator in Eq. (80) contains bT[{σ^z_p}], a conditional translation of Majoranas along domain-wall loops of arbitrary perimeter, and no circuit decomposition of this operator is provided. For a single closed loop of length L this is a cyclic Majorana shift, which is not manifestly a constant-depth nearest-neighbor circuit, so the theorem from [56] is doing essential work for both the definition of n2 in Sec. 5.1 and the obstruction proof in Sec. 6.1.1. The authors should state the precise theorem, supply a proof or a published reference, and verify that it applies to the conditional loop translation in Eq. (80). Without this, the central claim that a continuum-trivial Z_4^F symmetry forbids an SRE state is not established.","section":"Sec. 5.1, Eq. (41) and Sec. 7.1, Eq. (80)"},{"comment":"The proof that a nontrivial complex-fermion layer n3 obstructs SRE uses the statement 'Because n2 = 0, the 1D state |Φ>_{g;∂A} ... becomes a trivial invertible phase, hence SRE.' From the previous subsection, however, n2 = 0 only gives δχ = 0 for the cochain χ(g) defined by the Kitaev-chain class of |Φ>_{g;∂A}; χ can be a nontrivial element of H^1(G_b, Z_2), e.g. for G_b = Z_2 one may have χ(g) = 1. To reach the conclusion one must use the freedom to multiply U_A(g) by a boundary FQCA to shift χ by an arbitrary 1-cochain, a gauge freedom that is not stated in Sec. 5.1 or in Sec. 6.1. The gap is local and easily repaired, but it should be made explicit because it is needed for the reduction from the n3 layer to the bosonic layer.","section":"Sec. 6.1.2, after Eq. (69)"}],"minor_comments":[{"comment":"The phrase 'trivial up to circuit multiplications' is ambiguous; if every 2D fermionic QCA is a circuit, the 'up to' qualifier should be removed or explained.","section":"Sec. 5.1"},{"comment":"The notation [(-1)^{F_I}]^{ω2(g,h)} should use parentheses around the exponent to avoid ambiguity, for instance [(-1)^{F_I}]^{ω2(g,h)}.","section":"Sec. 2, Eq. (14)"},{"comment":"There is a typo in 'quantum cellular autamata' which should read 'quantum cellular automata'.","section":"Sec. 5.1"},{"comment":"Reference [56] is listed only as 'J. Haah, To appear.'; if a preprint or published version exists, it should be cited with full details, since the present version is not verifiable.","section":"References"},{"comment":"The statement that the continuum Z_4^F anomaly is trivial is supported by [20,58]; it would help the reader if the precise equivalence relation that identifies n2 = ω2 with the trivial class were quoted or derived.","section":"Sec. 7.2"}],"recommendation":"major_revision","confidential_remarks":"The main technical risk is the unpublished theorem [56]. If the authors cannot provide a proof or a verifiable reference, the central 2+1D example may be ill-posed, and the mismatch conclusion would not be supported. I do not see circularity or misrepresentation; the issue is a missing foundation for a key step. The manuscript is otherwise careful and the mismatch claim is interesting, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is worth your time. It gives a lattice-level characterization of fermionic 't Hooft anomalies by explicit index layers in (1+1)D and (2+1)D, and it constructs a concrete Z_4^F symmetry in (2+1)D that obstructs symmetric SRE despite having trivial continuum anomaly. The 1+1D part is largely a careful reworking of Else-Nayak and Seifnashri's results, but the proof that the H^1(BG_b,Z_2) layer is invisible to exact lattice symmetries is genuinely new, as is the odd-mod-8 no-go for Z_2. The 2+1D three-layer index (n2,n3,nu4) is new, and the explicit example is provocative.\n\nThe math in Secs. 2, 5, and 6 is internally consistent. The cocycle computations and gauge transformations check out; the stacking rule for (n2,nu3) is a nice compact formula. The obstruction-to-SRE arguments for the Majorana and complex-fermion layers are clear and don't smuggle in the conclusion. Credit where due: the paper flags its own main conjecture (onsiteability in 2+1D) and the absence of a converse.\n\nThe soft spot is structural and it is exactly the one the stress-test note identifies. The entire 2+1D construction assumes, via the unpublished reference [56], that every fermionic QCA in 2D is trivial up to circuits, so every finite internal symmetry is generated by an FDQC. That assumption is load-bearing: the truncation to a disk, the boundary QCA Omega_{dA}, and the index n2 are only defined for FDQC. The explicit Z_4^F generator in Eq. (80) contains bT, a controlled Majorana translation along arbitrary-length domain-wall loops. That operator is not manifestly an FDQC—for a loop of length L, the restriction to a disk seems to require depth O(L). If [56] does not cover this exact class of QCAs, then U_g is not FDQC by the paper's own criterion, and n2=omega_2 is not defined. The paper does not attempt to prove that bT is a circuit, and the unpublished reference cannot be inspected. That is a real gap, not a nitpick.\n\nThe 1+1D part does not suffer from this problem; the FDQC claim for finite internal symmetries in 1D is standard, and the onsiteability proof stands on its own.\n\nWho should read this: people tracking lattice vs continuum anomalies, fermionic SPTs, or QCA-based index constructions. It deserves a serious referee. I'd recommend sending it to review, but the referee must ask for one of two things: either a proof that the specific U_g in Eq. (80) is FDQC (or a modification of the example), or a version of the argument that defines n2 without relying on the unpublished theorem. The paper's core idea is sound, but the headline example currently rests on an unverifiable reference.\n\nBest","headline":"A solid lattice-anomaly paper with a genuinely interesting mismatch example, but the 2+1D result leans on an unpublished Haah theorem and the explicit Z_4^F operator may not be a finite-depth circuit.","tokens_in":879,"tokens_out":3773,"would_cite":true,"duration_ms":68239,"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 develops a lattice-level characterization of fermionic 't Hooft anomalies and shows that a lattice symmetry with trivial continuum anomaly can still forbid a symmetric short-range-entangled state.","keywords":["fermionic 't Hooft anomalies","lattice symmetries","short-range-entangled states","fermionic quantum cellular automata","Majorana translation","anomaly indices","central extension by fermion parity","tensor-product Hilbert spaces"],"falsifier":"A concrete falsifier would be an explicit symmetric short-range-entangled ground state for the $\\mathbb Z_4^F$ lattice symmetry of Sec. 7, which the paper's Majorana-layer index $n_2=\\omega_2$ says cannot exist.","tokens_in":21264,"feed_emoji":"⚛️","tokens_out":13805,"duration_ms":114990,"temperature":0.7,"pith_summary":"This paper develops a lattice-level characterization of fermionic 't Hooft anomalies for finite internal symmetries, phrased as obstructions to symmetric short-range-entangled states. In one spatial dimension an exact lattice symmetry carries a pair of cohomological indices $(n_2,\\nu_3)$; in two dimensions it carries a triple $(n_2,n_3,\\nu_4)$. For a symmetry whose bosonic group and fermion parity factorize, the paper proves that vanishing indices imply the symmetry can be rewritten as a strictly local product of single-site operations, hence admits a symmetric short-range-entangled state, while any nonvanishing index obstructs such a state. In the non-factorized case, a lattice index can obstruct a symmetric short-range-entangled state even when the continuum 't Hooft anomaly is trivial, as shown by an explicit $\\mathbb Z_4^F$ symmetry with $U_g^2=(-1)^F$ and Majorana-layer index $n_2=\\omega_2$.","feed_headline":"A lattice symmetry with trivial QFT anomaly can forbid a trivial state","feed_subtitle":"A symmetry whose square is fermion parity blocks any symmetric short-range-entangled state even though its continuum anomaly is zero","key_machinery":"The machinery is the truncation of the symmetry operator to a finite region and the study of the boundary operators that make the truncated group law hold. In one dimension the fermion parity of the left endpoint operator defines $n_2$, and the phase violation defines $\\nu_3$. In two dimensions the boundary operator's fermionic quantum-cellular-automaton index, taking values in $\\mathbb Q_+$ or $\\sqrt 2\\,\\mathbb Q_+$ according to whether it contains a Majorana translation, defines the Majorana layer $n_2$; after that layer is trivialized, an interval truncation of the boundary circuit defines the complex-fermion layer $n_3$ and then the bosonic layer $\\nu_4$. In the explicit example the symmetry generator is a domain-wall-dependent translation of Majorana fermions by one lattice site whose square is fermion parity, producing $n_2=\\omega_2$.","core_discovery":"The central discovery is that anomaly data extracted from an exact lattice symmetry do not have to match the anomaly classification of continuum quantum field theory. The paper obtains its indices by truncating a finite-depth-circuit symmetry operator to a disk or interval and reading the fermion parity or fermionic quantum-cellular-automaton index of the boundary operators that restore the group law. In (1+1)D, for $G_f=G_b\\times\\mathbb Z_2^F$, trivial $(n_2,\\nu_3)$ implies the symmetry is onsiteable, so the pair faithfully detects the lattice anomaly; but the continuum layer $H^1(BG_b,\\mathbb Z_2)$ cannot be realized by an exact lattice symmetry, and a $\\mathbb Z_2$ symmetry realizes only the even $\\mathbb Z_4$ subgroup of the continuum $\\mathbb Z_8$ classification. In (2+1)D each nonzero layer of $(n_2,n_3,\\nu_4)$ obstructs symmetric short-range-entangled states, and for a factorized symmetry it obstructs all symmetric invertible states. The central example is a $\\mathbb Z_4^F$ lattice symmetry with $U_g^2=(-1)^F$ and $n_2=\\omega_2$: it forbids a symmetric short-range-entangled state, yet its continuum 't Hooft anomaly is trivial because $n_2=\\omega_2$ is equivalent to zero in the continuum classification.","pith_inferences":["The same truncation machinery should generalize to (3+1)D, where boundary operators on a two-dimensional surface carry fermionic quantum-cellular-automaton indices; a four-layer hierarchy analogous to $(n_2,n_3,\\nu_4)$ would likely exhibit further lattice-continuum mismatches.","A direct test of the paper's picture is to build the predicted symmetric chiral invertible state with $c_-=1$ for the $\\mathbb Z_4^F$ symmetry; success would confirm that the lattice obstruction blocks short-range-entangled states but not all gapped symmetric states.","One can read the unattainable continuum layers, namely the $H^1$ layer in (1+1)D and the $p+ip$ layer under time reversal in (2+1)D, as a general rule: anomaly classes that require spatial translation or long-range entanglement in their microscopic realization cannot be exact internal symmetries of tensor-product lattices and must instead appear as emanant symmetries."],"forward_implications":["In (1+1)D, for $G_f=G_b\\times\\mathbb Z_2^F$, any exact symmetry with $(n_2,\\nu_3)=(0,0)$ is onsiteable and therefore admits a symmetric short-range-entangled state.","Continuum anomalies with a nonzero $H^1(BG_b,\\mathbb Z_2)$ layer, such as the odd elements of the $\\mathbb Z_8$ classification of $\\mathbb Z_2$ symmetry in (1+1)D, cannot be realized as exact internal symmetries on a tensor-product lattice; four copies of any exact $\\mathbb Z_2$ symmetry carry trivial lattice anomaly.","In (2+1)D, a nonzero value of any of $n_2,n_3,\\nu_4$ obstructs a symmetric short-range-entangled state, and when $G_f=G_b\\times\\mathbb Z_2^F$ it obstructs every symmetric invertible state.","The $\\mathbb Z_4^F$ symmetry with $U_g^2=(-1)^F$ and $n_2=\\omega_2$ forbids a symmetric short-range-entangled state while having a trivial continuum anomaly, so exact lattice symmetries can be incompatible with symmetric trivial phases for reasons invisible to continuum QFT.","The paper conjectures that in (2+1)D triviality of all three indices is also sufficient for onsiteability; if that conjecture holds, the triple is a faithful lattice anomaly classification, and anti-unitary anomalies with only a $p+ip$ layer would be unrealizable by exact symmetries."],"supporting_citations":[{"why":"introduces the fermion-parity defect index in one dimension that the paper extends to the pair (n2, nu3).","marker":"[1]"},{"why":"supplies the finite-depth-circuit disentangling construction used to prove onsiteability of trivial-index symmetries in (1+1)D.","marker":"[6]"},{"why":"provides the unpublished theorem that fermionic quantum cellular automata in two spatial dimensions are trivial up to circuit multiplications, on which the two-dimensional index construction rests.","marker":"[56]"},{"why":"establishes the sqrt-2-valued fermionic quantum-cellular-automaton index of Majorana translation used to define the Majorana layer n2.","marker":"[54]"},{"why":"fixes the cohomological anomaly data for continuum fermionic symmetry-protected phases, used for the lattice-continuum comparison.","marker":"[20]"},{"why":"classifies one-dimensional fermionic invertible states by Z2, used in the proof that a nonzero Majorana layer obstructs symmetric short-range-entangled states.","marker":"[57]"},{"why":"constructs the domain-wall Majorana-translation symmetry operator that the paper modifies to produce the Z4^F symmetry with n2 = omega2.","marker":"[7]"},{"why":"defines the bosonic anomaly layer nu4 and its obstruction to symmetric short-range-entangled states, which the paper adapts to the fermionic setup.","marker":"[4]"}],"fun_headline_variants":["Lattice vs continuum: trivial QFT anomaly still forbids SRE","Z4F symmetry: lattice anomaly blocks trivial states despite trivial QFT","Fermionic lattice anomaly that QFT misses","When a symmetry with no QFT anomaly forbids a trivial state","Exact lattice symmetries reveal extra anomalies beyond QFT"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that every finite internal fermionic symmetry on a two-dimensional tensor-product lattice can be written as a finite-depth quantum circuit up to a product of local circuits, a fact the paper cites to an unpublished theorem; if that fact fails, the boundary operators carrying the anomaly indices are not guaranteed to exist.","fun_headline_variants_meta":{"raw":{"variants":["Lattice vs continuum: trivial QFT anomaly still forbids SRE","Z4F symmetry: lattice anomaly blocks trivial states despite trivial QFT","Fermionic lattice anomaly that QFT misses","When a symmetry with no QFT anomaly forbids a trivial state","Exact lattice symmetries reveal extra anomalies beyond QFT"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000259,"raw_usage":{"total_tokens":1747,"prompt_tokens":1268,"completion_tokens":479,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":884,"completion_tokens_details":{"reasoning_tokens":392}},"tokens_in":884,"tokens_out":479,"duration_ms":4191,"temperature":1.0,"reasoning_tokens":392,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:09:00.077975+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete falsifier would be an explicit symmetric short-range-entangled ground state for the $\\mathbb Z_4^F$ lattice symmetry of Sec. 7, which the paper's Majorana-layer index $n_2=\\omega_2$ says cannot exist.","supporting_citations":[{"cited_title":"Classifying symmetry-protected topological phases through the anomalous action of the symmetry on the edge,","cited_arxiv_id":null,"evidence_quote":"introduces the fermion-parity defect index in one dimension that the paper extends to the pair (n2, nu3)."},{"cited_title":"To appear","cited_arxiv_id":null,"evidence_quote":"provides the unpublished theorem that fermionic quantum cellular automata in two spatial dimensions are trivial up to circuit multiplications, on which the two-dimensional index construction rests."},{"cited_title":"Interacting invariants for floquet phases of fermions in two dimensions,","cited_arxiv_id":null,"evidence_quote":"establishes the sqrt-2-valued fermionic quantum-cellular-automaton index of Majorana translation used to define the Majorana layer n2."},{"cited_title":"Construction and classification of symmetry-protected topological phases in interacting fermion systems,","cited_arxiv_id":null,"evidence_quote":"fixes the cohomological anomaly data for continuum fermionic symmetry-protected phases, used for the lattice-continuum comparison."},{"cited_title":"Unpaired majorana fermions in quantum wires,","cited_arxiv_id":null,"evidence_quote":"classifies one-dimensional fermionic invertible states by Z2, used in the proof that a nonzero Majorana layer obstructs symmetric short-range-entangled states."}],"review_version":1}