{"id":"85070db1-c11d-43ce-a2f6-8a146a485459","arxiv_id":"2412.11956","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"For the 2D massive Dirac equation in a constant magnetic field, the paper establishes microlocalized L1-to-Linfty decay of the form 2^{2j}(1+2^j t)^{-1/2} and local-in-time Strichartz estimates.","lead":"This paper proves new time-decay and Strichartz estimates for the massive Dirac equation in a constant magnetic field on the plane. The estimates are local in time because the magnetic field traps particles, and they provide tools for studying nonlinear Dirac equations in a uniform field.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The m=0 lowest-Landau-level case violates Proposition 2.8's τηλδα>0 hypothesis; at τηλδα=0 the claimed decomposition's remainder cannot decay, so the proof of Theorem 1.4 for m≥0 is incomplete.","rationale":"The reader's weakest assumption identifies Proposition 2.8 as the load-bearing step, and I agree. My analysis sharpens the concern: the failure is not merely a missing proof but a genuine obstruction at the spectral point x=B0, m=0, because the claimed rapidly decaying remainder cannot represent the non-decaying zero-mode propagator while the companion oscillatory integral decays by nonstationary phase. This directly affects Theorem 1.4, which states the decay estimate for all m≥0. Even if Lemma 2.10 were proved and supplied, Proposition 2.8 would still not apply to the lowest Landau level for m>0, nor to the zero mode for m=0, so the proof of (3.1)–(3.2) is incomplete as written. The intermediate-frequency gap in the passage from (3.1) to (3.2) is secondary because it can likely be patched by absorbing finitely many frequencies into C_T, but the zero-mode failure is structural. I therefore see no reason to alter the reader's REJECT verdict; the manuscript cannot be accepted without a self-contained proof of the decomposition and an explicit treatment of the lowest Landau level for m≥0.","tokens_in":17920,"tokens_out":18876,"duration_ms":155671,"concrete_test":"Provide a complete proof of Lemma 2.10, or locate a published proof, and check the decomposition (2.21) at the lowest Landau level. Concretely: fix m=0, x=B0, τηλδα=0, let f be a lowest-Landau-level eigenfunction with 2^{−j}√B0∈[1/2,1], and insert τηλδα=0 into (2.24). The exact left-hand side is ϕ(2^{−j}√H_{B0})f, which is independent of t. Bound the second term (2^jt)^{1/2}∫ χ(s,2^jt)e^{i2^jt/(4s)}ds by direct nonstationary phase; if it decays as (2^jt)^{−N} while the left-hand side is a fixed nonzero vector, then ρ must violate (2.22), proving Proposition 2.8 false as stated. If instead ρ is not rapidly decaying, identify the missing term and provide a separate argument for the zero mode that still yields (1.25) for m=0.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (1.25) for all m≥0 rests on Proposition 2.8, the microlocal decomposition of the half-wave propagator. This proposition is imported from the unpublished preprint [44], and its proof in §2.5 invokes Lemma 2.10 ('[44, Lemma]') without proof. More seriously, the hypotheses x>B0 and τηλδα=x+m²∓B0>0 exclude the lowest Landau level x=B0. At m=0, for the spin-up operator τηλδα=H_{B0}−B0, the eigenvalue τηλδα=0 occurs. At τηλδα=0 the left side of (2.21) is the non-decaying projection ϕ(2^{−j}√H_{B0}) (since e^{it√0}=1), while the second term on the right is the oscillatory integral (2^jt)^{1/2}∫ χ(s,2^jt)e^{i2^jt/(4s)}ds, whose phase 2^jt/(4s) has no critical point on supp χ⊂[1/16,8]. By nonstationary phase this term is O((2^jt)^{−N}) for any N. Hence the remainder ρ would have to be non-decaying, contradicting (2.22). Thus Proposition 2.8 cannot hold, or even apply, at the spectral point needed for m=0, and the paper gives no separate treatment of the zero mode. Even for m>0, x>B0 excludes the lowest Landau level, so the decomposition does not cover the full support of ϕ in Case 2 of Proposition 3.1. The proof of (3.2) also contains an incorrect frequency split ('for j≤j0, 2^jt≲1' fails for intermediate j), but the decisive gap is the unproved and hypothesis-violating stationary-phase lemma.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies local-in-time microlocal L^1-to-L^infty decay estimates and Strichartz estimates for the two-dimensional massive Dirac equation in a constant magnetic field. The main theorem asserts, for m≥0, the estimate (1.25) for the microlocalized propagator phi(2^{-j} sqrt(H_B0)) e^{itD_A}, and for m>0 the local Strichartz estimates (1.26). The proof reduces the squared Dirac operator to two Klein-Gordon-type operators H_B0+m^2∓B0, then uses a microlocal decomposition of the half-wave propagator (Proposition 2.8, taken from the unpublished preprint [44]), Bernstein inequalities adapted to the Landau Hamiltonian, and the known dispersive estimate for the Schrödinger propagator in a constant magnetic field.","tokens_in":18286,"tokens_out":9709,"duration_ms":81941,"significance":"If established, the result would be a useful contribution to dispersive PDE with unbounded magnetic potentials: the local-in-time microlocal decay (1.25) and the Strichartz estimates for m>0 would extend the known theory for Dirac equations in electromagnetic fields. The paper contains a clean algebraic reduction from the Dirac flow to two Klein-Gordon equations, and it correctly identifies the role of the Landau-level structure and the need for Bernstein-type inequalities. However, the m=0 case is false as stated, and the proof of the high-frequency representation depends on an unpublished lemma; these issues are load-bearing and prevent the paper from being accepted in its current form.","major_comments":[{"comment":"The estimate (1.25) cannot hold for m=0. Let f=(f_+,0) where f_+ is a lowest Landau level eigenfunction of H_B0 with eigenvalue B0, chosen so that f is in L1. For m=0, D_A^2 acts as H_B0-B0 on the upper component, so D_A^2 f=0 and hence D_A f=0; the propagator satisfies e^{itD_A}f=f for all t. Taking j with 2^j ~ B0^{1/2}, the left side of (1.25) is constant in t, while the right side contains the factor (1+2^j t)^{-1/2}, which decays. Thus (1.25) is contradicted. The failure is already visible in Proposition 2.8: its hypotheses x>B0 and tilde{x}=x+m^2-B0>0 exclude the spectral point x=B0 at m=0, where the half-wave phase is identically 1 and the oscillatory integral in (2.21) has no stationary phase. The m=0 case therefore requires a separate treatment or must be removed from the statement.","section":"Theorem 1.4, Eq. (1.25)"},{"comment":"The central decomposition (2.21) is imported from the unpublished preprint [44], and the proof in §2.5 cites Lemma 2.10, which is stated only as '[44, Lemma]' and is not proved in this paper. This is load-bearing: the whole of Case 2 in Proposition 3.1 depends on (2.21) and the estimates (2.22)-(2.23). Moreover, the hypothesis x>B0 excludes the lowest Landau level x=B0 even when m>0, where tilde{x}=m^2>0 but x>B0 fails. Since the support of phi(2^{-j} sqrt(x)) in the proof of Proposition 3.1 can intersect x=B0, the decomposition is not justified on the full spectral support of the microlocalization. The paper needs a proof of Lemma 2.10, a version of Proposition 2.8 valid for x≥B0 (or a separate handling of the lowest Landau level), and a clear statement of which parts of the proof depend on unpublished work.","section":"Proposition 2.8 and Lemma 2.10"},{"comment":"The passage from (3.1) to (3.2) contains an incorrect frequency split. The text says that for j≤j0, 2^j t ≲ 1, but the choice 2^{-j0}T ≤ π/(8B0) only implies 2^{j0} ≥ (8B0/π)T; for j close to j0 and t close to T, 2^j t can be large, so the first case does not apply. In Case 2, the estimate of the rho term requires |(m^2∓B0)/2^{2j}| ≤ 1/100, but the case condition t2^j ≫1 together with t≤T does not imply a lower bound on 2^j; for small j and large t the ratio (m^2∓B0)/2^{2j} need not be small, so the asserted bound 1/8 ≤ (λ_{k,ℓ}+m^2∓B0)/2^{2j} ≤ 8 is not guaranteed. Thus the proof does not establish the claimed finite-T estimate (3.2).","section":"Proposition 3.1, last paragraph"},{"comment":"The abstract states the decay estimate with the microlocalization phi(2^{-j}|D_A|), while Theorem 1.4 proves an estimate with phi(2^{-j} sqrt(H_B0)). Since |D_A| = diag(sqrt(H_B0+m^2-B0), sqrt(H_B0+m^2+B0)), these are different cutoffs for m>0; the two operators do not commute in the same way with the scalar spectral parameter. The claim advertised in the abstract is therefore not the claim proved in the body, and the abstract estimate is unsupported.","section":"Abstract versus Theorem 1.4"}],"minor_comments":[{"comment":"The manuscript contains several typos and stylistic issues, including 'deacy' in the title/abstract, 'filed' in §1, and 'Stricartz' in the reference list; these should be corrected.","section":"Throughout"},{"comment":"The displayed computation for ||D_A f||^2 ends with the expression ||nabla_A f_-||^2 - B0||f_-||^2 + ||nabla_A f_+||^2 + B0||f_+||^2, but the claimed bound by ||f||_{[H^1_A]^2} does not follow from this line without further argument, since the signs of the B0 terms differ. The proof should be clarified or the statement checked.","section":"Lemma 2.4"},{"comment":"The paper relies on [7] ('In preparation', with the author) for self-adjointness and on [44] for the key Lemma 2.10; both are unpublished. This makes the verification of the main proof difficult and should be disclosed explicitly in the introduction.","section":"References"}],"recommendation":"reject","confidential_remarks":"The m=0 counterexample in Theorem 1.4 is a concrete obstruction: the lowest Landau level gives a time-independent propagator, so the claimed decay cannot hold. Beyond the false m=0 statement, the proof for m>0 depends on an unpublished lemma and contains the frequency-split error described in the major comments. These are not merely presentation issues; the central claim as stated is not supported. Rejection is appropriate unless the authors substantially revise the statement and supply the missing arguments."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is the first paper I know to state microlocal decay and Strichartz estimates for the Dirac flow in a uniform magnetic field, and the reduction to a pair of Klein–Gordon equations via squaring D_A is the right way to start. But the main theorem, as stated for m≥0, is not proved. The load-bearing tool, Proposition 2.8, is imported from an unpublished preprint [44], and its hypotheses exclude the spectral point that matters for the m=0 spin-up component: x>B0 and x̃=x+m²∓B0>0 fail at the lowest Landau level x=B0. At x̃=0 the left side of (2.21) is time-independent, while the oscillatory integral has no stationary point on supp χ and decays rapidly; the remainder ρ would have to be non-decaying, contradicting (2.22). So the decomposition cannot hold, or even apply, where Theorem 1.4 needs it. Even for m>0, the condition x>B0 leaves out x=B0, so Case 2 of Proposition 3.1 does not cover the full support of φ. The paper contains no separate zero-mode argument.\n\nThere are smaller problems. The abstract uses φ(2^{-j}|D_A|) while Theorem 1.4 proves the scalar localization φ(2^{-j}√H_B0); these are not the same object. The proof of (3.2) contains a frequency split that does not work: 'for j≤j0, 2^j t≲1' is false, since j0 is chosen with 2^{-j0}T small, which makes 2^{j0}T large. Self-adjointness is deferred to [7] (in preparation), and the stationary-phase lemma to [44] (a preprint); that citation burden is noticeable but not fatal if the cited statements are true.\n\nWhat the paper does well: the squaring step is clean, the heat-kernel and subordination framework is the natural toolbox, and the Strichartz reduction via Keel–Tao is standard. The result, if fixed, would be a modest but real contribution: first dispersive bounds for Dirac in a constant field, useful for comparison with the nonrelativistic model and for nonlinear applications. The m>0 case away from the lowest Landau level looks salvageable; the fix is to prove the half-wave decomposition at x=B0, or to split off the zero mode and treat it by a separate argument.\n\nWho is this for: people working on dispersive estimates for magnetic operators, especially Dirac and Pauli equations. It deserves a serious referee: the gap is specific and possibly fixable, and the question is worth answering. I would not accept the current version, but I would send it to review rather than desk-reject it.","headline":"First claim of microlocal decay/Strichartz for Dirac in a uniform magnetic field, but the main theorem is not proved as stated because the imported half-wave decomposition fails at the lowest Landau level.","tokens_in":18844,"tokens_out":6949,"would_cite":false,"duration_ms":57981,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42B37","35Q40"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes microlocal L1-to-L∞ decay for the massive Dirac flow in a uniform magnetic field and derives local Strichartz estimates.","keywords":["decay estimates","Strichartz estimates","Dirac equation","constant magnetic field","Landau Hamiltonian","microlocal analysis","Klein-Gordon propagator","Littlewood-Paley theory"],"falsifier":"Using the explicit heat kernel (2.16), write the integral kernel of $\\varphi(2^{-j}\\sqrt{H_{B_0}})e^{it\\sqrt{H_{B_0}+m^2\\mp B_0}}$ in terms of the spectral eigenfunctions and evaluate its $L^1\\to L^\\infty$ operator norm at $t=2^{-j}$ for a sequence of $j$. The claim (1.25) predicts a uniform bound $C_T 2^{2j}(1+2^j|t|)^{-1/2}$; a single frequency $j$ where this norm grows faster than $2^{2j}$ would refute the theorem, while checking the $m=0$ lowest Landau level, where the $\\tilde{x}>0$ hypothesis of Proposition 2.8 is violated, would settle whether the massless range is valid.","tokens_in":17679,"feed_emoji":"🧲","tokens_out":10949,"duration_ms":92630,"temperature":0.7,"pith_summary":"The paper sets out to prove that solutions of the two-dimensional massive Dirac equation in a constant magnetic field decay at the same microlocal rate as a wave propagator: after localizing to frequencies $2^j$, the $L^\\infty$ norm decays like $2^{2j}(1+2^j|t|)^{-1/2}$ times the $L^1$ norm, for every finite time interval. This is presented as the first dispersive and Strichartz theory for a Dirac operator with an unbounded magnetic potential, a setting where perturbative methods break down. The proof exploits the spinorial structure by squaring the Dirac operator, which separates the two spinor components into Klein-Gordon equations governed by Pauli-type Landau Hamiltonians. As a consequence, the paper obtains local-in-time Strichartz estimates for the massive case $m>0$. The finite-time restriction is expected because the constant magnetic field traps particles, so global dispersion cannot hold.","feed_headline":"Uniform-field Dirac flow obeys wave-type decay estimates","feed_subtitle":"Local L1-to-L∞ decay and Strichartz estimates hold for massive Dirac equations, despite an unbounded trapping potential.","key_machinery":"The load-bearing identity is the square of the magnetic Dirac operator, $D_A^2=\\operatorname{diag}(H_{B_0}+m^2-B_0,\\,H_{B_0}+m^2+B_0)$, where $H_{B_0}$ is the Landau Hamiltonian, the magnetic Laplacian $(i\\nabla+A)^2$. The technical engine is a microlocal decomposition of the half-wave propagator, Proposition 2.8, which splits $\\varphi(2^{-j}\\sqrt{H_{B_0}})e^{it\\sqrt{H_{B_0}+m^2\\mp B_0}}$ into a rapidly decaying remainder plus an integral of Schr\\\"odinger propagators with effective time $2^{-j}t s$ and $s$ supported in $[1/16,8]$. Around this sit the explicit heat kernel (2.16), Bernstein inequalities for the spectral projectors of $H_{B_0}$, a square-function inequality, and an abstract Strichartz criterion that converts the decay estimate into the final $L^q_t L^p_x$ bound.","core_discovery":"The central claim is Theorem 1.4: with $A(x)=\\frac{B_0}{2}(-x_2,x_1)$ and $H_{B_0}=(i\\nabla+A)^2$, if $\\varphi(2^{-j}\\sqrt{H_{B_0}})f$ lies in $[L^1(\\mathbb{R}^2)]^2$, then for every finite $T$ there is a constant $C_T$ such that $$\\|\\varphi($2^{{-j}}$\\sqrt{H_{B_0}})$e^{{itD_A}}$f\\|_{[L^\\infty(\\mathbb{R}^2)]^2} \\le C_T $2^{{2j}}$(1+2^j|t|)^{-1/2} \\|\\varphi($2^{{-j}}$\\sqrt{H_{B_0}})f\\|_{[$L^{1}$(\\mathbb{R}^2)]^2},\\quad |t|\\le T.$$ The paper further proves the local Strichartz estimate (1.26) for $m>0$. The route is to square the Dirac operator, obtaining $D_A^2=\\operatorname{diag}(H_{B_0}+m^2-B_0,\\,H_{B_0}+m^2+B_0)$, which decouples the spinor evolution into two Klein-Gordon equations whose Hamiltonians differ by the Pauli shift $\\mp B_0$; the problem then reduces to proving wave-type microlocal decay for the half-wave propagators $e^{it\\sqrt{H_{B_0}+m^2\\mp B_0}}$.","pith_inferences":["A direct calculation on the lowest Landau level would test whether the $m=0$ range of Theorem 1.4 survives, since at that level the component $H_{B_0}+B_0$ has eigenvalue zero and Proposition 2.8's $\\tilde{x}>0$ hypothesis fails; this is a check the paper leaves open.","The same subordination-plus-projection strategy should transfer to other operators whose squares are Pauli-type shifts of a Landau Hamiltonian, such as three-dimensional Dirac operators in uniform fields, provided the corresponding Bernstein and square-function inequalities hold.","The local Strichartz estimates put nonlinear Dirac equations in uniform magnetic fields within reach of standard well-posedness arguments, since the admissible range now matches the wave equation's."],"forward_implications":["For every finite $T$ and every frequency $j$, the microlocalized Dirac flow decays like $C_T2^{2j}(1+2^j|t|)^{-1/2}$, matching the wave-propagator law.","For $m>0$, the local Strichartz estimate (1.26) holds for all admissible pairs $(q,p)\\in\\Lambda^W_s$ with $0\\le s\\le 1$.","The spinorial structure is essential: squaring $D_A$ diagonalizes the problem into two Klein-Gordon equations, so dispersive information for the Dirac flow is inherited from the Landau half-wave propagators.","Because the constant magnetic field is trapping, the estimates are local in time and the constant $C_T$ depends on the time horizon; global decay is not claimed."],"supporting_citations":[{"why":"Supplies Proposition 2.8, the microlocal decomposition of the half-wave propagator that the long-time high-frequency estimate depends on.","marker":"[44]"},{"why":"Supplies the $L^1\\to L^\\infty$ dispersive bound for the Schr\\\"odinger propagator $e^{itH_{B_0}}$ in a uniform magnetic field used in Case 2.","marker":"[32]"},{"why":"Supplies the abstract Strichartz criterion that turns the decay estimate into the frequency-summed Strichartz bound.","marker":"[31]"},{"why":"Supplies the spectral data and Besov/heat-kernel machinery for the Landau Hamiltonian used throughout Section 2.","marker":"[43]"},{"why":"Supplies the Rademacher-function argument for the square-function inequality associated with the spectral projectors of $H_{B_0}$.","marker":"[42]"},{"why":"Supplies the subordination-formula method for expressing half-wave propagators through Schr\\\"odinger propagators.","marker":"[16]"},{"why":"Supplies the explicit heat kernel (2.16) for $e^{-tH_{B_0}}$.","marker":"[41]"}],"fun_headline_variants":["Dirac decay in constant B: wave-type bounds","Strichartz for massive Dirac in magnetic field","Uniform field Dirac flow obeys wave decay","Massive Dirac in B-field gets Strichartz estimates","Wave-type decay proven for massive Dirac in constant B"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's load-bearing premise is Proposition 2.8, an imported, unproved decomposition of the microlocalized half-wave propagator from an unpublished companion preprint, on which the entire high-frequency long-time case depends.","fun_headline_variants_meta":{"raw":{"variants":["Dirac decay in constant B: wave-type bounds","Strichartz for massive Dirac in magnetic field","Uniform field Dirac flow obeys wave decay","Massive Dirac in B-field gets Strichartz estimates","Wave-type decay proven for massive Dirac in constant B"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000541,"raw_usage":{"total_tokens":2710,"prompt_tokens":1182,"completion_tokens":1528,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":798,"completion_tokens_details":{"reasoning_tokens":1454}},"tokens_in":798,"tokens_out":1528,"duration_ms":11230,"temperature":1.0,"reasoning_tokens":1454,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:27:16.036300+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Using the explicit heat kernel (2.16), write the integral kernel of $\\varphi(2^{-j}\\sqrt{H_{B_0}})e^{it\\sqrt{H_{B_0}+m^2\\mp B_0}}$ in terms of the spectral eigenfunctions and evaluate its $L^1\\to L^\\infty$ operator norm at $t=2^{-j}$ for a sequence of $j$. The claim (1.25) predicts a uniform bound $C_T 2^{2j}(1+2^j|t|)^{-1/2}$; a single frequency $j$ where this norm grows faster than $2^{2j}$ would refute the theorem, while checking the $m=0$ lowest Landau level, where the $\\tilde{x}>0$ hypothesis of Proposition 2.8 is violated, would settle whether the massless range is valid.","supporting_citations":[{"cited_title":"Decay estimates for one Aharonov-Bohm solenoid in a uniform magnetic field II: wave equation","cited_arxiv_id":"2309.07649","evidence_quote":"Supplies Proposition 2.8, the microlocal decomposition of the half-wave propagator that the long-time high-frequency estimate depends on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the $L^1\\to L^\\infty$ dispersive bound for the Schr\\\"odinger propagator $e^{itH_{B_0}}$ in a uniform magnetic field used in Case 2."},{"cited_title":"Keel and T","cited_arxiv_id":null,"evidence_quote":"Supplies the abstract Strichartz criterion that turns the decay estimate into the frequency-summed Strichartz bound."},{"cited_title":"Decay estimates for one Aharonov-Bohm solenoid in a uniform magnetic field I: Schr\\\"odinger equation","cited_arxiv_id":"2309.07635","evidence_quote":"Supplies the spectral data and Besov/heat-kernel machinery for the Landau Hamiltonian used throughout Section 2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Rademacher-function argument for the square-function inequality associated with the spectral projectors of $H_{B_0}$."},{"cited_title":"D’Ancona, V","cited_arxiv_id":null,"evidence_quote":"Supplies the subordination-formula method for expressing half-wave propagators through Schr\\\"odinger propagators."},{"cited_title":"Simon, Functional integration and quantum physics , Pure Appl","cited_arxiv_id":null,"evidence_quote":"Supplies the explicit heat kernel (2.16) for $e^{-tH_{B_0}}$."}],"review_version":1}