{"id":"e35a6347-54c2-4ea4-8c41-9722c01a2116","arxiv_id":"2608.13534","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Magnetic flows on negatively curved surfaces with nonpositive magnetic curvature retain topological transitivity, dense periodic orbits, and equilibrium states, but the natural magnetic distance is not symmetric and fails the triangle inequality.","lead":"This paper studies magnetic flows, paths of charged particles on curved surfaces, at the precise boundary where the system stops being uniformly chaotic. It shows that even at this edge, key dynamical features survive, and it reveals that the natural notion of distance fails in an interesting way.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reversed length/distance inequality and inverted β in Theorem 3.78 leave kinematic-expansivity, and hence equilibrium states, unproven.","rationale":"The reader's rationale identifies the reversed inequality in Theorem 3.78, and I agree that this is the central defect. However, the formal 'weakest_assumption' in the reader verdict points to the external Adachi theorem (Theorem 3.10); my concern is an internal proof failure in the main theorem, not the external premise. The flaw is load-bearing because Proposition 3.79 and Theorem 3.80 are immediate corollaries of Theorem 3.78. I also note the paper's positive side: it has many secondary results, a careful discussion of limitations (Remarks 4.11, 4.17), and an honest admission that magnetic horosphere theory is incomplete. But the proof of kinematic-expansivity contains a metric inequality in the wrong direction and an inconsistent use of β, so the submitted version does not meet the accept bar. The defect is local and plausibly repairable, so the rejection is not a dismissal of the program.","tokens_in":60065,"tokens_out":16104,"duration_ms":169567,"concrete_test":"Independently re-derive the flat-strip separation step of Theorem 3.78 in the model strip Γ(c, t) = (c, t + μ c t) with y_c ≡ 1 and x_c(t) = μt + c. For fixed large t, compute the Riemannian distance d(Γ(0, t), Γ(R, t)) and compare it with ∫_0^R |x_c(t)| dc = R(μt + R/2). If the distance is not bounded below by the integral, the proof of (39) fails; determine whether a weaker lower bound suffices. Also recheck the β-step: with β as defined, verify whether d(γ_v(t), γ_w(t)) ≥ β^{-1}(ε) follows from |t| ≥ ε, or whether the correct bound is β(ε).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 3.78 is the load-bearing step for entropy-expansivity (Prop. 3.79) and equilibrium states (Thm. 3.80). In the flat-strip case, the proof asserts d(γ_v(t), γ_w(t)) ≥ ∫_0^R |x_r(t)| dr > δ because (18) makes x_r(t) grow linearly. This reverses the metric inequality: for the curve c ↦ Γ(c, t), the length is ∫√(x_r² + y_r²) dr ≥ ∫|x_r| dr, and the Riemannian distance between the endpoints is at most that length, not at least. No argument shows the c-level curves are minimizing geodesics; in fact the orthogonality/normalization used in Lemma 3.55 silently drops the x_r-term, so the shear does not, by the proof given, separate points in the Knieper metric. The proof also misuses β: β maps magnetic distance to Riemannian distance, so β^{-1}(ε) is a time, and comparing the Riemannian distance d_K(w, v) with β^{-1}(ε) is dimensionally inconsistent. The theorem may be repairable, but as written the central claim is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies unit-speed magnetic (twisted geodesic) flows on closed negatively curved surfaces under the assumption (H1) of nonpositive magnetic curvature. It constructs a magnetic boundary homeomorphic to the Riemannian boundary, proves a magnetic visibility theorem, characterizes magnetically flat surfaces, develops a theory of magnetic flat strips, proves topological transitivity and density of periodic orbits, establishes orbit-equivalence to the geodesic flow under (H3), and claims kinematic-expansivity, entropy-expansivity of the time-one map, and existence of equilibrium states. A later part of the paper analyzes the failure of the magnetic triangle inequality and defines magnetic B-functions and orthospheres, proving their C^2 regularity. The central dynamical applications, especially the existence of equilibrium states, rest on Theorem 3.78, whose proof is the main point of concern.","tokens_in":60276,"tokens_out":10657,"duration_ms":123323,"significance":"If the proof gap in Theorem 3.78 is repaired, the paper would be a substantial contribution: it extends several nonpositive-curvature tools—boundary at infinity, visibility, flat-strip rigidity, and equilibrium states—to the magnetic setting, and it explicitly identifies the shear/twist of the magnetic flow as the mechanism producing kinematic expansivity. The paper is careful and extensive, with precise statements of limitations and open questions (Conjecture 3.36, Remarks 4.11 and 4.17), and it builds on cited external results rather than on fitted parameters. The main obstruction is the proof of the kinematic-expansivity theorem, which currently contains an invalid inequality and an inconsistent use of the function beta; since Proposition 3.79 and Theorem 3.80 depend on it, the advertised ergodic-theoretic conclusions are not established as written.","major_comments":[{"comment":"The proof's separation lower bound is invalid. For the cross-section c ↦ Γ(c,t), the tangent vector is the Jacobi field J_r(t) = x_r(t)γ'_r(t) + y_r(t)iγ'_r(t), so the length of the cross-section is ∫_0^R sqrt(x_r(t)^2 + y_r(t)^2) dr, which is at least ∫_0^R |x_r(t)| dr. The Riemannian distance between γ_v(t) and γ_w(t) is at most the length of this cross-section, not at least, because distance is the infimum of lengths over all curves. Thus the displayed inequality d(γ_v(t),γ_w(t)) ≥ ∫_0^R |x_r(t)| dr reverses the standard length-distance bound, and no argument is given that the cross-section is minimizing or that the coefficient x_r defines a 1-Lipschitz coordinate on the strip. Since (39) is the only mechanism producing δ-separation in the flat-strip case, Theorem 3.78, and with it Proposition 3.79 and Theorem 3.80, are not established as written.","section":"§3.7.2, Theorem 3.78, Eq. (39)"},{"comment":"The use of β in the final paragraph is order-inconsistent. Since β(l) = inf{d(p,q) : d̃(p,q) = l}, β^{-1}(ε) is a magnetic travel time, whereas d_K(w,v) is a Riemannian distance. For w = f_t v, the available lower bound is d(γ_v(0),γ_v(t)) ≥ β(t) because this pair has magnetic distance t, so |t| ≥ ε yields d ≥ β(ε), not d ≥ β^{-1}(ε). Monotonicity of β does not imply β(t) ≥ β^{-1}(ε) for t ≥ ε; indeed β(l) ≤ l, so β^{-1}(ε) can be much larger than ε. Consequently the choice δ = ½β^{-1}(ε) does not yield the claimed conclusion that |t| < ε. This is a second independent gap in the same theorem.","section":"§3.7.2, Theorem 3.78, definition of β and final paragraph"}],"minor_comments":[{"comment":"Theorem 3.10 is quoted from [Ada97] without stating its hypotheses; since Corollary 3.11 underpins the magnetic distance, the boundary construction, visibility, and flat-strip theory, please state the exact hypotheses (including any compactness or regularity conditions) and give a precise reference, or include a proof sketch.","section":"§3.1, Theorem 3.10 and Corollary 3.11"},{"comment":"The displayed estimate 'd(γ_δ(r_δ(t)), γ_0(t) ≤ ...' is missing a closing parenthesis, and the family of reparametrizations r_s should be defined before it is used.","section":"§3.2, Proposition 3.26 proof"},{"comment":"In the estimate near (55), the expression 'µ(x_1 − x_2) + (˙y_1 − ˙x_2)' appears to contain a typo and should presumably read 'µ(x_1 − x_2) + (˙y_1 − ˙y_2)'.","section":"§4.3, Proposition 4.19 proof"},{"comment":"In item (5) of the proof, the notation 'byξ' is undefined and probably should be an arc such as 'cyξ'; please clarify.","section":"§3.4, Theorem 3.42 proof"},{"comment":"The function β is described as continuous and increasing, but the proof only needs a nondecreasing function with the stated separation property; please state whether β is strictly increasing or, if not, explain how the inverse β^{-1} is defined.","section":"§3.7.2, Theorem 3.78, Eq. (38)"}],"recommendation":"major_revision","confidential_remarks":"The invalid inequality in Theorem 3.78 is the main obstruction. The claimed theorem is plausible and the gap is localized to the flat-strip separation argument, so I recommend major revision rather than rejection: if the authors supply a correct lower bound for d(γ_v(t),γ_w(t)) in the flat-strip case—for example by exhibiting a suitable 1-Lipschitz coordinate along the strip—the paper would be a strong candidate for publication. The rest of the geometric development is careful and substantial."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the quick take: this is a serious, ambitious paper that develops a coherent geometric framework for magnetic flows on negatively curved surfaces when magnetic curvature is nonpositive. The magnetic boundary homeomorphism, the flatness characterization, visibility under (H2), the shear/twist structure of magnetic flat strips, and the orbit-equivalence criterion are new and mostly well argued. The reader's report is right, though: Theorem 3.78 does not go through as written, and since it supports entropy-expansivity and the existence of equilibrium states, the paper's central claim is not established.\n\nThe core problem is a metric inequality that runs backward. In the flat-strip case the proof asserts d(γ_v(t), γ_w(t)) ≥ ∫_0^R |x_r(t)| dr. But the curve c ↦ Γ(c,t) has length ∫ sqrt(x_r² + y_r²) dr, and the Riemannian distance between its endpoints is at most that length, not at least. No argument shows the c-level curves are minimizing geodesics; the orthogonality/normalization in Lemma 3.55 actually drops the tangential term, so the shear does not, on the proof given, separate points in the Knieper metric. There is also a dimensional slip: β maps magnetic distance to Riemannian distance, so β^{-1}(ε) is a time, and comparing d_K(w,v) with it doesn't make sense. These are local, repairable problems, but they break the load-bearing argument for kinematic-expansivity.\n\nEverything around it—the boundary construction, the flat-strip dynamics, topological transitivity, density of periodic orbits, orbit-equivalence—looks solid or at least carefully argued. The paper is also honest about what it cannot yet do (magnetic horospheres, continuity in the boundary point; see Remarks 4.11 and 4.17). The reliance on Adachi's theorem is external but standard.\n\nWho should read it? Anyone working on magnetic flows, nonuniform hyperbolicity, or equilibrium states. The geometric program is valuable and will likely be cited once the proof is fixed. It deserves a serious referee, but the submitted version shouldn't be accepted: the author needs to repair Theorem 3.78.","headline":"Substantial geometric program for magnetic flows at the threshold of hyperbolicity, but the key kinematic-expansivity theorem has a proof error that leaves the equilibrium-state result unproven.","tokens_in":60813,"tokens_out":3010,"would_cite":false,"duration_ms":32052,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37D40","37D35"],"pacs":[],"model":"deepseek-v4-flash","headline":"On negatively curved surfaces, magnetic flows remain kinematic-expansive and admit equilibrium states even when magnetic curvature is merely nonpositive; orbit-equivalence to the geodesic flow holds exactly when no magnetically flat strip…","keywords":["magnetic flows","twisted geodesic flows","hyperbolic dynamics","nonpositive magnetic curvature","equilibrium states","magnetic boundary","kinematic-expansivity","orthospheres"],"falsifier":"Search for a closed negatively curved surface and a value of $\\mu$ with $K+\\mu^2\\le 0$ and $K+\\mu^2<0$ somewhere for which the $\\mu$-magnetic exponential map fails to be a covering map, for example by numerically finding a nontrivial $\\mu$-magnetic Jacobi field whose orthogonal component $y(t)$ vanishes at two distinct times; such a magnetic conjugate pair would break the quoted covering-map theorem and with it the uniqueness of magnetic geodesics on which the paper's boundary and flat-strip results rest.","tokens_in":59834,"feed_emoji":"🧲","tokens_out":7307,"duration_ms":69444,"temperature":0.7,"pith_summary":"The paper studies magnetic (twisted geodesic) flows on closed negatively curved surfaces at the edge of hyperbolicity, where the magnetic curvature $K+\\mu^2$ is allowed to vanish somewhere while the underlying surface curvature stays negative. It sets out to show that the main dynamical conclusions familiar from geodesic flows survive under this weaker, nonuniform hyperbolicity: a magnetic boundary at infinity, visibility, dense periodic orbits, topological transitivity, and the existence of equilibrium states. The central surprise is that shearing along magnetically flat strips separates orbits linearly in time, so the flow remains kinematic-expansive even where ordinary expansivity and orbit-equivalence to the geodesic flow fail. The result is a clean dichotomy: ordinary expansivity and orbit-equivalence hold exactly in the absence of magnetically flat strips, while kinematic-expansivity, entropy-expansivity, and equilibrium states hold for every such flow.","feed_headline":"Flat magnetic strips still allow equilibrium states","feed_subtitle":"Even at zero magnetic curvature, the flow shears orbits apart and still has a measure of maximal entropy.","key_machinery":"The engine is the $\\mu$-magnetic Jacobi equation for the orthogonal component, $\\ddot{y}+(K+\\mu^2)y=0$, coupled to the tangential component by $\\dot{x}=\\mu y$. Under $K_\\mu\\le 0$ the function $y$ is convex, which provides comparison bounds, stability of asymptotic magnetic geodesics, and the dichotomy between exponential separation and flat parallel Jacobi fields. In a magnetically flat strip the equations integrate to the shear matrix $Df^\\mu_t=\\begin{pmatrix}1&\\mu t\\\\0&1\\end{pmatrix}$, which separates orbits linearly in time. The Knieper metric $d_K(v,w)=\\max_{t\\in[0,1]}d(\\gamma_v(t),\\gamma_w(t))$ converts that shear into kinematic-expansivity, and hence into entropy-expansivity of the time-1 map. Magnetic orthospheres, defined as curves orthogonal to families of asymptotic magnetic geodesics, supply the $C^2$ regularity needed for strong (un)stable spaces on the regular set.","core_discovery":"For a closed surface with $K<0$ and magnetic parameter $\\mu$ satisfying $K_\\mu=K+\\mu^2\\le 0$, the $\\mu$-magnetic flow is weakly hyperbolic rather than uniformly hyperbolic. The paper establishes that each such flow has a magnetic boundary homeomorphic to the usual ideal boundary, that any two distinct boundary points are joined by a magnetic geodesic whenever $K_\\mu<0$ somewhere, and that multiple connecting magnetic geodesics occur exactly along magnetically flat Euclidean strips filled with singular orbits. It proves that uniqueness of connecting magnetic geodesics (condition (H3)) is equivalent both to ordinary expansivity and to orbit-equivalence with the underlying geodesic flow. Without that uniqueness, shearing along flat strips destroys expansivity but still separates orbits, so the flow is kinematic-expansive and its time-1 map is entropy-expansive; consequently every Bowen-bounded potential with finite pressure has an equilibrium state, including a measure of maximal entropy. The paper also shows the natural magnetic distance is asymmetric and satisfies only a partial magnetic triangle inequality, and it develops magnetic orthospheres as $C^2$ substitutes for horospheres, yielding strong stable and unstable spaces on regular vectors.","pith_inferences":["Editorial inference: if the covering-map premise is valid, the shear picture suggests a magnetic translation length that varies across flat strips; measuring that variation would give a magnetic marked length spectrum testable by comparing periods of nearby magnetic geodesics.","Editorial inference: because entropy-expansivity comes from a bounded spanning set across flat strips, entropy and pressure computations for such flows may be reducible to local measurements inside strips, which is numerically accessible.","Editorial inference: the $C^2$ orthosphere construction opens the way to a Pesin-type stable manifold theory on the dense regular set; the open continuity question for orthospherical leaves as the boundary point varies is the natural next step toward uniqueness of equilibrium states.","Editorial inference: a direct numerical experiment integrating magnetic geodesics in a flat-strip region should show orthogonal separation growing linearly at rate $\\mu t$, matching the shear matrix, and exponential separation only where $K_\\mu<0$; this cleanly separates the weakly hyperbolic regime from the uniformly hyperbolic one."],"forward_implications":["The magnetic boundary at infinity is homeomorphic to the Riemannian ideal boundary, and under (H2) every pair of distinct boundary points is joined by a magnetic geodesic with distinct forward and backward endpoints.","Under (H2), periodic orbits are dense, regular periodic vectors are dense, and the flow is topologically transitive with nonwandering set equal to the entire unit tangent bundle.","Ordinary expansivity and orbit-equivalence to the geodesic flow each hold exactly under (H3), the absence of magnetically flat strips.","Even when flat strips are present, the time-1 map is entropy-expansive, so every Bowen-bounded potential with finite pressure has an equilibrium state, including a measure of maximal entropy.","The natural magnetic distance is not a metric: it is asymmetric, fails the triangle inequality on one side of a magnetic segment, and obeys only a partial triangle inequality; magnetic orthospheres are $C^2$ and support strong stable and unstable spaces on regular vectors."],"supporting_citations":[{"why":"Supplies the covering-map theorem for the $\\mu$-magnetic exponential map on which uniqueness of magnetic geodesics (Corollary 3.11) and the whole boundary construction rest.","marker":"[Ada97, Theorem 3]"},{"why":"Shows magnetic rays cross geodesic circles exactly once and are unbounded, giving the endpoint maps and the magnetic boundary-chord setup.","marker":"[Ada96, Theorem 1]"},{"why":"Provides the sharp angle bound and orbit-equivalence result for negative magnetic curvature that the paper extends to (H2) and (H3).","marker":"[Gro99b, Theorem 3.1]"},{"why":"Derives the $\\mu$-magnetic Jacobi equation and the variation formalism defining stable, unstable, and parallel magnetic Jacobi fields.","marker":"[PP96]"},{"why":"Supplies the visibility, cone-topology, and transitivity arguments for geodesic flows that the paper adapts to prove magnetic visibility and topological transitivity.","marker":"[EO73]"},{"why":"Gives density of endpoint pairs of axial isometry axes, used to prove density of periodic magnetic orbits.","marker":"[Bal82, Theorem 2.13]"},{"why":"Introduces the Knieper metric and the entropy-expansivity framework for time-1 maps of geodesic flows in nonpositive curvature.","marker":"[Kni98]"},{"why":"Gives the implication from kinematic-expansivity to entropy-expansivity of the time-1 map and the equilibrium-state construction applied in Theorem 3.80.","marker":"[CT16, Proposition 3.3]"},{"why":"Supplies the radial-flow and Sturm-comparison method used to prove $C^2$ regularity of magnetic orthospheres and strong (un)stable spaces.","marker":"[HIH77]"},{"why":"Provides the standard definitions of expansivity, orbit-equivalence, topological pressure, and equilibrium states used throughout the statements.","marker":"[FH19]"}],"fun_headline_variants":["Nonpositive magnetic curvature still yields maximal entropy measures","Flat magnetic strips don't kill equilibrium states","Weak hyperbolicity still gives entropy measures","No loss of equilibrium states in flat magnetic strips"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes without proof a theorem of Adachi: under $K<0$ and $K+\\mu^2\\le 0$, the $\\mu$-magnetic exponential map from any point of the universal cover is a covering map; if that premise fails, the uniqueness of magnetic geodesics and everything built on it collapses.","fun_headline_variants_meta":{"raw":{"variants":["Nonpositive magnetic curvature still yields maximal entropy measures","Flat magnetic strips don't kill equilibrium states","Weak hyperbolicity still gives entropy measures","No loss of equilibrium states in flat magnetic strips"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00068,"raw_usage":{"total_tokens":3014,"prompt_tokens":794,"completion_tokens":2220,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":410,"completion_tokens_details":{"reasoning_tokens":2163}},"tokens_in":410,"tokens_out":2220,"duration_ms":17455,"temperature":1.0,"reasoning_tokens":2163,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:47:26.230817+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a closed negatively curved surface and a value of $\\mu$ with $K+\\mu^2\\le 0$ and $K+\\mu^2<0$ somewhere for which the $\\mu$-magnetic exponential map fails to be a covering map, for example by numerically finding a nontrivial $\\mu$-magnetic Jacobi field whose orthogonal component $y(t)$ vanishes at two distinct times; such a magnetic conjugate pair would break the quoted covering-map theorem and with it the uniqueness of magnetic geodesics on which the paper's boundary and flat-strip results rest.","supporting_citations":[],"review_version":1}