{"id":"dcf9fc0b-fafe-4c6d-ba94-4e9ba287ac80","arxiv_id":"2411.17944","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Exact Gaussian fluctuation calculations for two linearly coupled non-reciprocal fields reveal enhanced k^-4 and k^-6 divergences at critical exceptional points and a mechanism for 1/f noise.","lead":"This paper solves exactly the simplest random, diffusion-coupled pair of non-reciprocal fields and shows where they become unstable, where fluctuations get anomalously strong, and how coupling one field to another can turn white noise into colored, 1/f-like noise. It gives researchers a clean linear benchmark for more complex non-equilibrium and active-matter theories.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'additional CEP' on the hyperbola for D1≠D2 is not an exceptional point: the k^-4 divergence there arises from a vanishing gradient coefficient, not eigenvector coalescence, so the paper overstates the occurrence of CEPs.","rationale":"The reader's weakest assumption concerns the use of equal-time correlations on the oscillatory critical line m1+m2=0. This is less dispositive than the CEP mislabeling because the equal-time correlation function is a well-defined observable even for oscillatory modes: the k^-2 divergence signals the marginal k=0 mode, and the paper explicitly distinguishes this line as having rotational dynamics. The static susceptibility at ω=0 is finite, but equal-time correlations still provide a valid, if non-equilibrium, diagnostic of criticality. By contrast, the claim that a point with D2 m1 + D1 m2 = 0 on the hyperbola is a CEP is internally inconsistent with the paper's own definition of a CEP as a point where exceptional lines meet critical lines. The exceptional line is (m1−m2)^2 + 4Δ = 0, which on the hyperbola is exactly m1 = −m2; for D1≠D2 the tangency point does not lie on that line. The k^-4 divergence is genuine, but it arises from a different mechanism (vanishing of the linear-in-k^2 coefficient in the denominator), and should not be presented as a critical exceptional point. This affects the central claim that the paper demonstrates CEPs with enhanced fluctuations, since one of the two claimed types of CEP is not exceptional. The correction is a matter of terminology and physical interpretation rather than a change in the exact correlation-function formulas, so the overall conditional-accept verdict remains appropriate, but the paper must be revised to remove the mischaracterization.","tokens_in":1218,"tokens_out":1057,"duration_ms":258351,"concrete_test":"Take D1=4, D2=1, Δ=−1, and choose the point on the hyperbola m1m2=Δ with D2 m1 + D1 m2 = 0: m1=2, m2=−1/2. The discriminant (m1−m2)^2 + 4Δ = (2.5)^2 − 4 = 2.25 > 0, so the eigenvalues of the mean-field matrix are 0 and −1.5, distinct and non-degenerate; the eigenvectors are linearly independent. This demonstrates the point is not an exceptional point. Repeating the same check at the true CEP m1=−m2 (e.g., m1=1, m2=−1, Δ=−1) gives discriminant zero, confirming eigenvalue coalescence there.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the Δ<0 regime, the paper labels every point on the critical hyperbola where D2 m1 + D1 m2 = 0 as a 'CEP' (Sec. II.B, around Eq. (9)). This condition makes the second bracket in Eq. (9) scale as k^2, giving a k^-4 denominator, but it is not the exceptional-point condition. The exceptional-point condition for the evolution matrix is (m1−m2)^2 + 4Δ = 0 (Eq. (2)). On the critical line m1m2 = Δ, this reduces to (m1+m2)^2 = 0, so true CEPs occur only at m1 = −m2. For generic D1≠D2, the point satisfying D2 m1 + D1 m2 = 0 has m1+m2 ≠ 0, so the eigenvalues are distinct: one is zero (critical) and the other is −(m1+m2) ≠ 0. The matrix is diagonalizable, so there is no exceptional point. The enhanced k^-4 divergence is real, but it is a momentum-space tangency (the mode-stability curve is tangent to the critical line), not a consequence of eigenvector coalescence. Attributing this to a CEP misidentifies the mechanism and may mislead predictions about nonlinear extensions, since true CEPs are known to host distinct RG physics.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the simplest linear, spatially extended, stochastic two-field models with non-reciprocal couplings, in both overdamped and inertial forms. It derives exact equal-time correlation functions, maps stability regions as a function of the non-reciprocity parameter Δ, and identifies divergence laws of the correlations near critical lines: the usual k^-2 divergence, enhanced k^-4 and k^-6 divergences at certain points labeled critical exceptional points (CEPs), and a finite-momentum instability for strong non-reciprocity and unequal diffusion constants. It further claims that integrating out one field generates 1/f-type noise for the other. The manuscript is written as a reference-level classification of linear non-reciprocal Gaussian models, with explicit formulas that are internally consistent in their main algebraic steps.","tokens_in":12106,"tokens_out":5102,"duration_ms":47319,"significance":"If the claims are correct, the paper provides a useful exact solvable baseline for perturbative and RG treatments of non-reciprocal field theories, similar in spirit to the Gaussian fixed point in equilibrium critical dynamics. Its strengths are that the correlation functions are derived explicitly without fitting parameters, the stability conditions and denominator structures are transparent, and the paper makes falsifiable predictions about the momentum divergence exponents at special parameter values. However, three load-bearing points currently undercut the conclusions as stated: the frequency exponent in the 1/f noise calculation has the wrong sign, the 'additional CEP' on the critical hyperbola is not an exceptional point, and the equal-time correlation diagnostic is used on an oscillatory critical line despite the paper's own caveat that this diagnostic may be inappropriate there. These are correctable in revision but affect the paper's central classification claims.","major_comments":[{"comment":"The evaluation of the integrated-out noise term is not correct as written. Equation (13) states that ∫ d^d k j12^2 B2/(ω^2 + (m2 + D2 k^2)^2) ∼ j12^2 B2 D2^{-d/2} ω^{(4-d)/2} for d ≤ 3 and ω >> m2. Dimensional analysis of the integral gives instead a prefactor D2^{-d/2} ω^{(d-4)/2}: for d=1 this is ω^{-3/2}, for d=2 it is ω^{-1}, and for d=3 it is ω^{-1/2}. Thus the claimed generic '1/f noise' does not follow; only in d=2 is the spectrum exactly 1/f. The exponent sign error is load-bearing for the paper's final conclusion about colored noise, and the text and the inertial analogue in Sec. III should be corrected accordingly.","section":"Sec. II.B, Eq. (13)"},{"comment":"The point on the critical hyperbola where D2 m1 + D1 m2 = 0 is called a 'CEP' (critical exceptional point), but it is not an exceptional point of the evolution matrix. The exceptional-point condition for Eq. (2) is (m1 − m2)^2 + 4Δ = 0; combined with the critical line m1 m2 = Δ this reduces to (m1 + m2)^2 = 0, so true CEPs have m1 = −m2. At the additional point with D2 m1 + D1 m2 = 0 and D1 ≠ D2, the eigenvalues are 0 and −(m1 + m2) with m1 + m2 ≠ 0, so the matrix is diagonalizable. The enhanced k^-4 divergence there is real but is a momentum-space tangency (the mode-stability line is tangent to the hyperbola), not eigenvector coalescence. This distinction matters because the paper uses CEP language to connect to known RG phenomenology; the text should rename or explicitly qualify these points and avoid claiming that all enhanced divergences are CEP effects.","section":"Sec. II.B, around Eqs. (9) and (11)"},{"comment":"The k^-2 divergence along the m1 + m2 = 0 line is computed from the static equal-time correlation function, even though the paper itself states in Sec. II.B that 'looking at the static correlation function might be inappropriate' for the oscillatory critical line in the Δ<0 regime. The manuscript does not reconcile this caveat with the subsequent claim that this line 'also exhibits the same divergence as the other critical lines.' If static correlations are not the correct diagnostic on this line, the claimed critical divergence there is not established. The authors should either justify the use of equal-time correlations in this regime or provide a frequency-dependent analysis, and should temper the classification of this line as a critical line with k^-2 behavior.","section":"Sec. II.B, around Eq. (10)"}],"minor_comments":[{"comment":"The frequency variable ω and the '1/f' notation are not defined consistently; if ω is angular frequency, the statement '1/f noise' should specify whether the power spectrum is ∝ 1/ω or 1/f, and the corrected exponents from Eq. (13) should be presented in a consistent notation.","section":"Throughout"},{"comment":"The heading 'INER TIAL SYSTEMS' contains a typo and should read 'INERTIAL SYSTEMS'.","section":"Section heading"},{"comment":"The caption for Fig. 1(c) says that 'the critical lines all have k^-2 divergence in the correlators,' but the text immediately discusses points with k^-4 and k^-6 divergences on those lines; the caption should be qualified to distinguish generic points from the special points.","section":"Fig. 1 caption"},{"comment":"The displayed denominator in Eq. (17) is typeset in a way that is hard to parse; adding explicit parentheses around the two brackets would improve readability.","section":"Eq. (17)"},{"comment":"Reference [26] is listed as 'See Supplemental Material' but the paper does not indicate an ancillary file or a full citation; the reference should be completed.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The central Gaussian calculation appears internally consistent and the paper addresses a useful gap, but the three major issues above are all load-bearing for the stated conclusions. The 1/f exponent sign error and the mischaracterization of the 'additional CEP' are easily fixed by a careful revision, and the oscillatory-line issue can be resolved by either a new diagnostic or a clear limitation. The paper is not beyond repair, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives the exact Gaussian correlators for a minimal linear non-reciprocal two-field model with white noise and diffusion. That is genuinely useful: it provides a clean linear benchmark for MSR calculations and shows k^-4 and k^-6 divergences at double-tuned points, which I don't think are in the earlier literature. The stability diagram in terms of Delta is neat, and the finite-momentum instability mechanism is clearly explained. The effective noise calculation after integrating out one field is also a nice addition. There is no fitting and no circularity here; the calculation is straightforward but done carefully.\n\nThe main problem is the labeling of the 'additional CEP' along the hyperbola for D1≠D2. The condition D2 m1 + D1 m2 = 0 makes the second bracket in Eq. (9) scale as k^2, giving a k^-4 denominator. But this is not an exceptional point. On the critical line m1m2 = Delta, the eigenvalue-coalescence condition reduces to (m1+m2)^2 = 0, so true CEPs sit at m1 = -m2. For generic D1≠D2, the tangency point satisfies m1+m2 ≠ 0, so the evolution matrix is diagonalizable with one zero eigenvalue and one nonzero. The enhanced divergence is real, but it is a momentum-space tangency—the line of modes with slope D2/D1 grazing the critical hyperbola—not eigenvector coalescence. The paper's own definition of CEP in Sec. I agrees with this reading, so calling the tangency point a CEP is internally inconsistent. This matters because true CEPs have distinct RG physics, and conflating the two could mislead nonlinear extensions. It is a fixable terminology/interpretation issue, but it should be fixed.\n\nTwo smaller things. The paper uses the equal-time correlator on the oscillatory critical line m1+m2 = 0 in the Delta<0 regime, while acknowledging that static correlations might be inappropriate there. That needs a better justification. And the '1/f noise' claim is dimension-dependent: the integrated noise goes as omega^{(4-d)/2}, so it is exactly 1/f only in d=2. The paper says '1/f type' in the abstract, which is fair, but the wording in the body is a bit loose.\n\nThe central calculation holds up. I think this deserves a serious referee; a revision that corrects the CEP terminology and clarifies the static-correlator caveat would make it a solid reference.","headline":"The exact Gaussian correlators and k^-6 divergence are new and useful, but the 'additional CEP' along the hyperbola for D1≠D2 is really a tangency effect, not an exceptional point.","tokens_in":12690,"tokens_out":4935,"would_cite":true,"duration_ms":39355,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["05.40.-a","64.60.Ht","05.70.Ln"],"model":"deepseek-v4-flash","headline":"The paper shows that in the simplest linear stochastic model of two non-reciprocally coupled fields, the equal-time correlation functions diverge as $k^{-2}$ on ordinary critical lines, as $k^{-4}$ at critical exceptional points, and up…","keywords":["non-reciprocal systems","Gaussian fluctuations","critical exceptional points","equal-time correlation functions","finite-momentum instability","1/f noise","reaction-diffusion systems","linear stochastic dynamics"],"falsifier":"Numerically integrate the overdamped equations (1) on a finite lattice in $d = 1$ for $\\Delta < 0$ with parameters placed on the critical line $m_1 = -m_2$, and measure the equal-time structure factor $S(k) = \\langle |\\phi_1(k,t)|^2 \\rangle$; if $S(k)$ does not scale as $k^{-2}$ while the frequency-resolved spectrum shows a divergence at a nonzero frequency, the static-criterion claim fails. Independently, at the $\\Delta = 0$ CEP $m_1 = m_2 = 0$, check the predicted $k^{-4}$ and $k^{-6}$ power laws of the exact expressions (8) by direct numerical solution; a mismatch in any exponent would refute the central enhancement claim.","tokens_in":2350,"feed_emoji":"🌀","tokens_out":4918,"duration_ms":111271,"temperature":0.7,"pith_summary":"The paper works out the exact Gaussian statistics of the simplest spatially extended, non-reciprocally coupled pair of fields: two linear stochastic reaction-diffusion equations with white noise and unequal couplings $j_{12} \\neq j_{21}$. Its central finding is that the non-reciprocity parameter $\\Delta = j_{12} j_{21}$ controls the entire stability portrait: it enlarges the stable region, creates exceptional lines where the two modes coalesce, and, where those lines meet a critical line, it sharpens the usual $k^{-2}$ divergence of the equal-time correlation function into $k^{-4}$ and, at isolated doubly tuned points, $k^{-6}$ divergences. The paper also shows that sufficiently strong non-reciprocity combined with unequal diffusion constants produces a finite-momentum instability, and that eliminating one field leaves the other with induced colored noise of 1/f type. Because the model is exactly solvable, it provides a linear reference point—analogous to the Gaussian model in equilibrium—for perturbative and renormalization-group studies of non-reciprocal systems.","feed_headline":"Non-reciprocal systems sharpen critical fluctuations to k^-6","feed_subtitle":"A solvable two-field model reveals where exceptional points meet critical lines and why 1/f noise appears.","key_machinery":"The load-bearing object is the two-field Gaussian action built through the Martin-Siggia-Rose–Janssen–De Dominicis procedure, from which the equal-time correlation functions (8) are computed exactly. The non-reciprocity parameter $\\Delta = j_{12} j_{21} = j_+^2 - j_-^2$ controls the stability condition $m_1 m_2 > \\Delta$ and organizes the phase diagram; because the coupling matrix is non-normal, exceptional points (where two eigenvalues and their eigenvectors coincide) appear whenever the argument of the square root in the dispersion (2) changes sign, and critical exceptional points are the intersections of those exceptional lines with critical lines. At such points the denominators of (8) gain extra powers of $k^2$, which is the entire mechanism behind the $k^{-4}$ and $k^{-6}$ enhancements. The same denominator analysis, applied to the momentum-shifted parameters $m_i + D_i k^2$, yields the finite-momentum instability, and the effective single-field action obtained by integrating out $\\phi_2$ supplies the noise-kernel replacement that generates 1/f noise.","core_discovery":"For the overdamped system (1), the exact equal-time correlation functions (8) show that along generic critical lines $m_1 m_2 = \\Delta$ the denominator yields the standard $k^{-2}$ behavior of equilibrium Model A dynamics, so mild non-reciprocity is statistically indistinguishable from equilibrium criticality. At a critical exceptional point, where the eigenvalues and eigenvectors of the dynamical matrix coalesce at the same time as the system is critical, the denominator acquires an extra vanishing factor and the correlations diverge as $k^{-4}$; when an exceptional point sits at the intersection of two critical lines (for example $m_1 = m_2 = 0$ at $\\Delta = 0$, or the corners in the $\\Delta < 0$ phase diagram with equal diffusion constants), the divergence reaches $k^{-6}$. In the strongly non-reciprocal regime $\\Delta < 0$, the stability region is expanded and an oscillatory region appears; along the critical line $m_1 + m_2 = 0$ the orbits are closed, and the correlation functions still show the same $k^{-2}$ divergence, though the paper states that the static diagnostic is questionable there. If the ratio of diffusion constants is large enough, the modes cross the critical line at finite momentum, giving a finite-momentum instability; at the two critical momenta the correlator diverges as $|k - k_c|^{-1}$, while tangency at a CEP gives $|k - k_c|^{-2}$. Integrating out one field replaces the noise kernel $B_1$ by $B_1 + j_{12}^2 B_2 / (\\omega^2 + (m_2 + D_2 k^2)^2)$, which in dimensions $d \\leq 3$ produces 1/f-type noise, and the inertial extension (14) preserves the same qualitative hierarchy of divergences.","pith_inferences":["If the enhanced CEP divergences persist after nonlinearities are added, the non-reciprocal Allen-Cahn model and similar nonlinear extensions would have a lower upper critical dimension than equilibrium Model A, a difference that could be tested by comparing fluctuation corrections in $d = 3$.","The noise-kernel replacement suggests a generic experimental signature: in any two-species activator-inhibitor or excitatory-inhibitory system near criticality, the power spectrum of one species should show a 1/f component inherited from the other species even when the couplings are reciprocal.","The paper leaves the frequency-resolved correlation function on the oscillatory critical line $m_1 = -m_2$ uncomputed; a natural extension is to look for a spectral rather than static divergence there, which would connect the transition to time-crystalline order discussed for non-reciprocal Ising models.","The $k^{-4}$ and $k^{-6}$ scalings at CEPs could serve as a practical probe in neural and active-matter experiments: spatial correlations decaying faster than the equilibrium expectation would indicate that the system sits at a critical exceptional point."],"forward_implications":["Perturbative and renormalization-group treatments that start from this linear theory must use the stronger $k^{-4}$ and $k^{-6}$ divergences at critical exceptional points to set their upper critical dimensions, not the equilibrium $k^{-2}$ value.","A system that is unstable under reciprocal couplings can be made stable by adding non-reciprocity, because the stability region in the $(m_1, m_2)$ plane widens as $\\Delta$ decreases.","Strong non-reciprocity with sufficiently different diffusion constants produces a finite-momentum instability; the resulting patterns have a preferred length scale set by $k_c$, and near onset the correlator divergence is $|k - k_c|^{-1}$ (or $|k - k_c|^{-2}$ at a tangency with a CEP).","Any linear field coupled to a second diffusive field picks up an additional noise term whose spectrum is 1/f-like in $d \\leq 3$; the 1/f tail extends to all frequencies as the system approaches a critical exceptional point.","The inertial generalization shows the same $\\Delta$-controlled stability diagram, the same $k^{-2}/k^{-4}/k^{-6}$ hierarchy of correlations, and an additional CEP on the internal critical line when $\\gamma_1 = \\gamma_2$."],"supporting_citations":[{"why":"Supplies the definition of the non-reciprocity parameter through symmetric and antisymmetric coupling parts and the broader framework of non-reciprocal phase transitions.","marker":"[1]"},{"why":"Establishes critical exceptional points in nonequilibrium O(N) models, giving the phenomenology the paper's CEPs are compared with.","marker":"[11]"},{"why":"Demonstrates critical fluctuations at a many-body exceptional point, providing the $k^{-4}$-type enhancement this paper generalizes.","marker":"[12]"},{"why":"Provides the field-theoretic methods and the equilibrium Model A $k^{-2}$ scaling used as the baseline throughout the paper.","marker":"[16]"},{"why":"Gives the reaction-diffusion and pattern-formation context that motivates the finite-momentum instability analysis.","marker":"[21]"},{"why":"Contains the phase plots and the detailed integrating-out calculation that yield the effective action and the noise spectrum.","marker":"[26]"},{"why":"Provides the nonlinear non-reciprocal Allen-Cahn extension whose pattern formation the paper's finite-momentum instability anticipates.","marker":"[27]"},{"why":"Shows noise amplification from non-orthogonal eigenvectors, the effect the induced 1/f noise is compared with.","marker":"[28]"}],"fun_headline_variants":["Non-reciprocity sharpens critical noise to k^-6","Exceptional points push fluctuations to k^-6","How non-reciprocity yields 1/f noise and k^-6 peaks","Finite-momentum instability from strong non-reciprocity","From equilibrium k^-2 to non-reciprocal k^-6 noise"],"cache_read_input_tokens":14720,"weakest_assumption_plain":"The classification of critical lines in the $\\Delta < 0$ regime rests on using the equal-time correlation function, even on the oscillatory line $m_1 = -m_2$ where the paper itself says the static correlation function may be an inappropriate diagnostic; if static correlations mischaracterize that transition, the $k^{-2}$ divergence claimed there is not the relevant critical quantity.","fun_headline_variants_meta":{"raw":{"variants":["Non-reciprocity sharpens critical noise to k^-6","Exceptional points push fluctuations to k^-6","How non-reciprocity yields 1/f noise and k^-6 peaks","Finite-momentum instability from strong non-reciprocity","From equilibrium k^-2 to non-reciprocal k^-6 noise"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00108,"raw_usage":{"total_tokens":4610,"prompt_tokens":1126,"completion_tokens":3484,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":742,"completion_tokens_details":{"reasoning_tokens":3393}},"tokens_in":742,"tokens_out":3484,"duration_ms":22513,"temperature":1.0,"reasoning_tokens":3393,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:41:47.990205+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Numerically integrate the overdamped equations (1) on a finite lattice in $d = 1$ for $\\Delta < 0$ with parameters placed on the critical line $m_1 = -m_2$, and measure the equal-time structure factor $S(k) = \\langle |\\phi_1(k,t)|^2 \\rangle$; if $S(k)$ does not scale as $k^{-2}$ while the frequency-resolved spectrum shows a divergence at a nonzero frequency, the static-criterion claim fails. Independently, at the $\\Delta = 0$ CEP $m_1 = m_2 = 0$, check the predicted $k^{-4}$ and $k^{-6}$ power laws of the exact expressions (8) by direct numerical solution; a mismatch in any exponent would refute the central enhancement claim.","supporting_citations":[{"cited_title":"Hanai and P","cited_arxiv_id":null,"evidence_quote":"Demonstrates critical fluctuations at a many-body exceptional point, providing the $k^{-4}$-type enhancement this paper generalizes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the field-theoretic methods and the equilibrium Model A $k^{-2}$ scaling used as the baseline throughout the paper."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the reaction-diffusion and pattern-formation context that motivates the finite-momentum instability analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the nonlinear non-reciprocal Allen-Cahn extension whose pattern formation the paper's finite-momentum instability anticipates."}],"review_version":1}