{"id":"5189d676-2dae-4dbb-862d-69a5d90109de","arxiv_id":"2607.14874","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The (1+1)D Dirac quantum cellular automaton has fermion-doubling poles that genuinely contribute to its Green's function; its one-step Green's function has a closed form, and the flavor-staggered fixed model's Green's function is built directly from the original model's chiral components.","lead":"This paper proves that a quantum cellular automaton version of the Dirac equation creates fake extra fermion modes, and it derives simple formulas for the one-step propagator of both the original and a \"flavored\" fixed version. It also shows when that automaton or ordinary lattice fermions better approximate the real Dirac equation away from the continuum limit.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"App. B's contour integral is ill-defined: the poles z± lie exactly on the integration contour, so the claimed non-vanishing contribution of every determinant zero is not established.","rationale":"The reader's weakest_assumption focuses on the witness criterion being confined to the free, one-time-step sector and not the full interacting or n-step model. My concern is more specific and located inside that witness computation: the residue integral that is supposed to prove every determinant zero contributes non-vanishingly is not mathematically well-defined, because the poles lie on the integration contour. This does not refute the central claim — the Dirac QCA very likely does exhibit fermion doubling, and Eq. (25) is recoverable as the one-step unitary propagator — but it does undermine the 'definitive proof' as written. The paper would need to specify a contour prescription (e.g., causal iη) and show the surviving pole contributions are prescription-independent, or alternatively prove the doubling claim by computing the n-step GF and identifying the spurious propagating modes directly. Since this is an addressable gap rather than a demonstrated counterexample, the reader's CONDITIONAL verdict remains appropriate. I therefore recommend no change to the verdict, but the revision should address the contour prescription explicitly.","tokens_in":34593,"tokens_out":17173,"duration_ms":166018,"concrete_test":"Repeat the energy integration in Eq. (B2) with an explicit causal deformation E → E + iη and separately E → E − iη, map to z, take η → 0⁺, and compare the resulting one-step GF with Eq. (25). If the retarded and advanced prescriptions differ from Eq. (25) (e.g., half-residue factors or only one pole contributing), the pole-contribution claim is prescription-dependent and the current derivation does not establish the fermion-doubling witness.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step for the 'definitive proof' is App. B's residue evaluation of the energy integral. The roots in Eq. (B14), z± = c_ε cos(pϵ) ± i sqrt(1 − c_ε² cos²(pϵ)), satisfy |z±| = 1 for every p because c_ε = cos(mϵ). Hence the poles of f(z; p) lie on the integration contour C, the unit circle, not inside it. The paper assigns Ind_C(z±) = −1 and applies the full residue formula 2πi Res; that is valid only for interior poles. A pole on the contour makes the integral ill-defined without an explicit prescription (principal value or an iη deformation pushing the poles off the circle). Different prescriptions give different coefficients — typically half-residue contributions or only one pole contributing — so the assertion that every zero of the determinant contributes non-vanishingly to the one-time-step GF is not a well-defined mathematical result as written. The closed form Eq. (25) is independently obtained in App. C as the quantum-walk propagator, so the GF expression itself is not necessarily wrong; what is unproven is the fermion-doubling witness, namely that the pole structure of G_B(E,p) is exactly what the one-step energy integral detects. This is precisely the kind of missing contour prescription that should be required before the word 'definitive' is warranted.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the (1+1)-dimensional Dirac QCA of Arrighi et al. It derives a closed form for the one-time-step Green's function, Eq. (25), by energy contour integration, and interprets the poles of the Fourier-space determinant as fermion doublers. It compares the QW integrand with the continuum Dirac propagator and with continuous-time naive LGT in ultrarelativistic and nonrelativistic regimes. It then computes the Green's function of the flavored Dirac QCA of Ref. [3], obtaining Eq. (65), which expresses the FQCA Green's function in terms of the original QW Green's-function blocks, with a flavor operator I or σ_x depending on the parity of D+Δn.","tokens_in":34944,"tokens_out":17725,"duration_ms":147755,"significance":"The closed-form one-step Green's function and the compact FQCA Green's-function expression are potentially useful results, and the comparison of lattice models in far-from-continuum regimes addresses a relevant quantum-simulation question. The combinatorial/inductive proof of Eq. (65) in Apps. J and K is a strength: it is self-contained and does not depend on the questionable Fourier normalization. The quantitative model comparison is also clearly presented. However, the 'definitive proof' of fermion doubling rests on a contour integration whose poles lie on the integration contour; as written, that part is not rigorous.","major_comments":[{"comment":"The roots in Eq. (B14) satisfy |z_±|^2 = c_ε^2 cos^2(pε) + 1 - c_ε^2 cos^2(pε) = 1, so both poles lie on the unit circle C, which is the integration contour. The residue theorem with Ind_C(z_±) = -1 and full 2πi residues is therefore not applicable: a simple pole on the contour makes the integral ill-defined unless a principal value or an iη deformation is specified, and different prescriptions give different coefficients. This invalidates the Sec. IIIC claim that every zero of D(E,p) contributes non-vanishingly to the one-time-step GF. Please add an explicit regularization (e.g., replace e^{-iEε} by e^{-iEε-η} in L_B) and redo the computation, or prove the pole-contribution claim by the direct unitary-propagator route and state the distributional identity under that prescription.","section":"App. B, Eqs. (B10)-(B15)"},{"comment":"There is a sign/prefactor inconsistency between Eq. (B7) and the final result: combining dE = i dz/(εz) with the integrand in Eq. (B4) yields an overall sign different from that in Eq. (B15), and Eq. (B15) appears to give minus Eq. (22). Since Eq. (25) is independently confirmed by the propagator computation in App. C, this is not fatal, but the contour-integral derivation must be corrected to match.","section":"App. B, Eqs. (B7), (B15)"},{"comment":"The rhombus-Brillouin-zone Fourier ansatz, including the factor 2=(√2)^2, is assumed rather than derived. The final FQCA result Eq. (65) is nevertheless proven by the independent combinatorial and inductive arguments in Apps. J and K, so the main result survives; however, the Fourier-integral derivation in Sec. IVA3 is incomplete unless the integration measure and normalization are justified.","section":"Sec. IV, Eq. (45)"},{"comment":"The paper claims a definitive proof of fermion doubling for a QCA model used as a foundation for QED, but the explicit computation is limited to the free single-particle sector and to one time step. This is a scope limitation rather than a technical error—the resolvent pole structure is the standard free-field diagnostic—but the title/abstract should either restrict the claim accordingly or add a sentence explaining why the one-step free GF suffices for the interacting-QFT conclusion.","section":"Sec. I and Sec. V"}],"minor_comments":[{"comment":"The imaginary part of I^QW_11 is missing the sin(εp) factor and incorrectly contains cos(εm) sin(εm) instead of cos(εm) sin(εp). Eq. (41d) likewise has a spurious e^{iPD}; both should read (1/2π) cos(M) sin(P) for -Im{e^{-iPD}F^QW_11}.","section":"App. H, Eq. (H3d); Sec. III.D, Eq. (41d)"},{"comment":"The manuscript contains many typos and grammatical errors (e.g., 'quantum cellular automatas', 'we use QCAs will be used', duplicated words). A careful proofread is needed.","section":"General"},{"comment":"The FD-fixing model and the FQCA construction are taken from the authors' own preprint Ref. [3]; please ensure it is published or otherwise available, and clarify which definitions are being imported.","section":"Ref. [3]"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely correct in its main closed-form results, but the 'definitive proof' framing is too strong given the App. B issue. The authors' earlier work Ref. [3] supplies the FQCA construction and the doubling identification; the present paper's novelty is the Green's-function computation. The title and abstract overstate generality (interacting QED) relative to the free 1+1D computation. I would recommend major revision rather than rejection because Eq. (25) and Eq. (65) are corroborated by independent derivations."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good to get your take on this. I've now spent some time with it, and I largely agree with your conditional read—with one thing sharper: the stress-test note is right. The contour integral in App. B is not a well-defined object as written. After the change of variable z = e^{-iεE}, the roots z± = c_ε cos(pε) ± i sqrt(1 − c_ε² cos²(pε)) lie exactly on the unit circle, which is the integration contour. You can't assign an index to a point on the contour, and the residue theorem doesn't apply without a prescription. If you push the contour outward you get both poles' residues; if you push it inward you get zero. The paper just states Ind_C(z±) = −1 and applies the full residue formula. That's invalid. So the advertised 'definitive proof' that every determinant zero contributes non-vanishingly is not established by App. B.\n\nThat said, the main closed-form Eq. (25) is not in doubt. App. C derives the same thing directly from the quantum-walk propagator, and App. D does the Fourier integral. The paper's real value is that closed form, plus the FQCA Green's function in Eq. (65), which is a clean result. The regime comparison with LGT is also thoughtful, and the plots are honest. The FQCA structure proofs in Apps. J and K are solid given the rhombus-BZ ansatz from Ref. [3]; the factor of 2 in Eq. (45) is assumed, not derived, so the FQCA result inherits that assumption.\n\nOther issues: the title says 'definitive proof' but the analysis is restricted to a free one-time-step sector; that's a paper-scope problem, not a fatal one. App. H has a typo in (H3d): the QW b term should be cos(εm) sin(εp), not sin(εm). And a couple of supporting claims are explicitly unpublished, which is fine but weakens the self-containedness.\n\nWho is this for? People working on QCA-based QED and quantum simulation of Dirac dynamics will want the closed-form one-step GF and the fixed-model GF. The fermion-doubling argument needs to be cleaned up before anyone can rely on it as a proof, but the paper is useful and worth reviewing. I'd send it to a referee, with the instruction that the contour prescription must be addressed—either by an iη deformation or by reframing the 'pole contribution' argument in terms of the direct propagator. If the authors fix that, the paper becomes a solid contribution.","headline":"The closed-form one-step GF and the FQCA GF are useful, but the advertised 'definitive proof' rests on an ill-defined contour integral—the poles lie on the contour—so the proof needs a prescription before it can be accepted.","tokens_in":35440,"tokens_out":5562,"would_cite":true,"duration_ms":47165,"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":"The Dirac QCA's propagator genuinely has fermion doubling, and the paper gives its closed-form Green's function.","keywords":["fermion doubling","Dirac quantum cellular automaton","discrete-time quantum walk","Green's function","lattice gauge theory","quantum electrodynamics","flavor staggering","Brillouin zone"],"falsifier":"A direct numerical check: initialize the Dirac quantum walk on a single spatial site and measure the one-step Green's function matrix elements for all four chirality combinations; Eq. (25) predicts exact coefficients c_ε at x-ε and x+ε and -i s_ε at x. If any measured amplitude differs, the closed-form propagator is wrong. For the doubling claim, compute the full n-step Green's function and look at the residue at a spurious pole such as the one near (E,p)=(π/(2ε), -π/(2ε)); if that residue vanishes for some n or for some boundary conditions, the assertion that every zero contributes non-vanish","tokens_in":34438,"feed_emoji":"⚛️","tokens_out":8290,"duration_ms":68294,"temperature":0.7,"pith_summary":"This paper tries to establish that the Dirac quantum cellular automaton—a unitary, chirality-dependent spacetime-lattice discretization of the Dirac equation used to build lattice QED models—really does suffer from fermion doubling, and not just in its dispersion relation. In 1+1 dimensions the authors prove the point by showing that every spurious pole of the Fourier-space Green's function contributes non-vanishingly to the one-time-step Green's function after contour integration. They also find that this one-step Green's function is a closed form built from Kronecker deltas and the sine and cosine of mε, much simpler than the Dirac equation's Green's function. For the flavor-staggered, doubling-free version of the model, they show that its Green's function is a parity-selected combination of the original model's chiral components. A sympathetic reader would care because the result identifies where spurious modes live in QCA-based QFT and provides exact two-point correlation functions to build on.","feed_headline":"Fermion doubling is real in the Dirac QCA propagator","feed_subtitle":"The one-step Green's function is an exact closed form; the flavor-staggered fix's correlator reduces to the original's chiral parts.","key_machinery":"The load-bearing objects are the Fourier-space equation-of-motion operator and its determinant D(E,p), whose zeros are the poles of the Green's function. The decisive move is the change of variable z=e^{-iεE}, which converts the energy integral over the Brillouin zone into a contour integral on the unit circle; the residue theorem then forces each pole to contribute with winding number -1, so the doublers cannot be discarded by a choice of contour. In direct space, the closed form of Eq. (25) is the exact one-step propagator, and it is what makes the doubling concrete. For the fixed model, the key machinery is the rhombus (diamond) Brillouin-zone ansatz with its factor of 2 and the decomposi","core_discovery":"The paper's central claim is that the free single-particle sector of the (1+1)D Dirac QCA—the Dirac discrete-time quantum walk with one-step unitary U_ε = [[cos(mε)S_ε, -i sin(mε)],[-i sin(mε), cos(mε)S_ε†]]—has a genuine fermion-doubling problem. The Fourier-space equation-of-motion determinant D(E,p) has extra zeros; writing the energy integral as a contour integral in z=e^{-iεE} and applying the residue theorem, the authors show that every such zero is a pole that contributes non-vanishingly to the direct-space one-time-step Green's function. That Green's function is the closed form G(t'+ε,x;t',x')=(1/ε)[[c_ε δ_{x',x-ε}, -i s_ε δ_{x',x}],[-i s_ε δ_{x',x}, c_ε δ_{x',x+ε}]]. For the flavore","pith_inferences":["Because the one-step Green's function is a closed form, the n-step Green's function likely satisfies a simple convolution recurrence; one could derive closed-form n-step propagators by repeated application of Eq. (25), which the paper leaves open.","The same contour-integral witness could be applied to the continuous-time LGT and to the 2+1 and 3+1 QCA models to settle whether their doublers also contribute non-vanishingly to their propagators, extending the proof beyond 1+1 dimensions.","The regime comparison suggests a practical guide for quantum simulators: choose the QCA realization for massless or very light fermions and the continuous-time LGT for massive, non-relativistic fermions; a hybrid scheme that switches in the intermediate regime might be worth testing, although the paper does not propose one.","The FQCA Green's function's parity selection rule is a sharp signature that an experiment on a two-flavor diamond-lattice walk could verify by preparing a single-site state and measuring the flavor-chirality correlations after an even or odd number of steps."],"forward_implications":["The one-time-step Green's function of the Dirac QCA is known in closed form, so single-step propagation of any initial state on the lattice can be done exactly and cheaply.","Fermion doubling is a property of the propagator itself, so any interacting QED-QCA built on this Dirac QCA inherits the spurious modes unless the flavor-staggering fix is applied.","In ultrarelativistic regimes (small mε) the Dirac QCA approximates the continuum Dirac equation better than continuous-time naive lattice fermions, while the opposite holds in non-relativistic regimes (mε near π/2).","The flavored QCA's two-point function is fully determined by the four chiral components of the original model's Green's function via a simple parity rule, giving an exact and computationally tractable correlation function for the FD-fixed model.","The paper claims the Dirac QCA's doubling is less severe by a factor of three than that of discrete-time standard lattice gauge theories, because its extra modes are spatiotemporal rather than separate spatial and temporal doublers."],"fun_headline_variants":["Dirac QCA has fermion doubling, proven","Fermion doubling proven in Dirac QCA","Green's function of Dirac QCA is simple closed form","Flavored QCA fix reduces to original's chiral parts"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof abstracts from the interacting QFT to the free single-particle Dirac quantum walk, so the load-bearing assumption is that every spurious pole of the free one-particle Green's function—and nothing else—fully determines fermion doubling in the interacting QED-QCA; if the energy-contour witness misses or over-counts modes once gauge fields are present, the 'definitive proof' would not transfer to the full QFT.","fun_headline_variants_meta":{"raw":{"variants":["Dirac QCA has fermion doubling, proven","Fermion doubling proven in Dirac QCA","Green's function of Dirac QCA is simple closed form","Flavored QCA fix reduces to original's chiral parts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000253,"raw_usage":{"total_tokens":1543,"prompt_tokens":1025,"completion_tokens":518,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":769,"completion_tokens_details":{"reasoning_tokens":462}},"tokens_in":769,"tokens_out":518,"duration_ms":4327,"temperature":1.0,"reasoning_tokens":462,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T00:48:12.866039+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct numerical check: initialize the Dirac quantum walk on a single spatial site and measure the one-step Green's function matrix elements for all four chirality combinations; Eq. (25) predicts exact coefficients c_ε at x-ε and x+ε and -i s_ε at x. If any measured amplitude differs, the closed-form propagator is wrong. For the doubling claim, compute the full n-step Green's function and look at the residue at a spurious pole such as the one near (E,p)=(π/(2ε), -π/(2ε)); if that residue vanishes for some n or for some boundary conditions, the assertion that every zero contributes non-vanish","supporting_citations":[],"review_version":1}