{"id":"6b783ca2-ad3b-4944-b932-4c4c21820982","arxiv_id":"2607.18891","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In a 1D bistable gene-toggle lattice, coarsening proceeds by abrupt cascades and shows log-periodic oscillations, with measured exponents δ<0.5, deviating from the Allen–Cahn value δ=1/2.","lead":"This paper simulates a one-dimensional line of diffusively coupled bistable gene switches and finds that domain walls freeze for long stretches, then vanish in sudden cascade events rather than moving smoothly. The result suggests that collective pattern reorganization in multicellular systems can be intermittent and hierarchically timed rather than a gradual smoothing process.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central DSI claim relies on an unvalidated exponential collapse law; direct cascade-time ratios must be measured before λ_t≈3.476 can be accepted.","rationale":"I read the paper in good faith. The qualitative picture—pinned fronts, abrupt cascade annihilation, and log-periodic decay—is internally plausible and supported by the representative spatiotemporal plots. The quantitative claim, however, depends on a specific mechanism: exponentially growing collapse times as a function of integer domain length, plus a roughly one-site growth of disappearing domains per cascade. The authors themselves flag the exponential law as heuristic and non-quantitative, and no direct measurement of successive cascade-time ratios is reported. This is exactly the load-bearing weak point identified by the Reader: if the exponential collapse law fails quantitatively, the constant geometric spacing λ_t≈3.476 is not established, and the DSI interpretation reduces to a log-periodic fit without a mechanistic explanation. The reader's CONDITIONAL verdict is appropriate: the paper should provide direct cascade-time data, ensemble error bars, a principled D_c determination, and a quantitative test of Eq. (7). My stress-test does not change that verdict, so I mark UNCHANGED. I also note the internal inconsistency of the δω≈2 relation across Table 1, which reinforces the need for more careful quantitative reporting. No ad hominem is intended; the concern is about the strength of the evidence relative to the claim.","tokens_in":9879,"tokens_out":5786,"duration_ms":54726,"concrete_test":"Using the stored (or rerun) trajectories for α=10 at D_c=0.092071296, extract the actual times t_i at which ρ(t) jumps (cascades), e.g., by thresholding the derivative of the smoothed ρ(t), and compute the successive ratios t_{i+1}/t_i for all cascades over t∈[0,2.5×10^6]. If the mean/median ratio is not constant and within uncertainty of 3.476—and does not match e^{2π/ω}=3.465—across at least 5 independent cascades/runs, then the geometric-spacing claim underlying temporal DSI is not supported. Repeat for the other D_c values in Table 1 and report ensemble error bars.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing premise is the quantitative discrete-scale-invariance picture in Sections 5–6. The argument is: (i) domain collapse time grows exponentially with domain length, t_c ∼ e^{κL}/(κ v_0) (Eq. 8); (ii) disappearing-domain lengths grow roughly linearly with cascade index, L_i ∼ i (Fig. 5); (iii) therefore successive cascade times are geometrically spaced, yielding λ_t = t_{i+1}/t_i ≈ 3.476 for α=10, consistent with the log-periodic fit λ = e^{2π/ω} ≈ 3.465. The weak point is (i), which is not quantitatively established. Appendix A.2–A.4 derives the exponential law by linearizing the reaction term around the single-cell low fixed point, and the authors explicitly concede that κ is 'only qualitative' and 'does not provide a quantitative prediction of cascade times.' Moreover, Section 5 does not report a direct measurement of the sequence of cascade times t_i; the quoted λ_t appears to be inferred from the assumed exponential law combined with the measured L_i. If the prefactor v_0 varies with L, or if the front interaction is not a pure exponential over the relevant range of L, then λ_t ≈ 3.476 is not an independent observable and the DSI interpretation loses its quantitative footing. In addition, the supporting δω≈2 relation is not satisfied by Table 1 (δω ≈ 1.98, 2.38, 2.38, 1.70), further weakening the 'inverse scaling' suggestion. The qualitative cascade phenomenology and the presence of log-periodic oscillations are plausible, but the central DSI claim as a quantitative statement is underdetermined.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies deterministic coarsening in a one-dimensional lattice of diffusively coupled bistable gene-toggle switches. It reports that domain walls remain pinned for long intervals and then disappear through collective cascade events, that the domain-wall density decays as ρ(t)∼t^{-δ} with δ<0.5 for all promoter strengths α considered, and that the decay has superimposed log-periodic oscillations. The authors argue that these oscillations are controlled by the sizes of domains disappearing in each cascade: because collapse times grow exponentially with domain length (Eq. 7) and the average disappearing-domain length grows roughly linearly with cascade index, successive cascade times are geometrically spaced, giving λ_t≈3.476 for α=10, consistent with temporal discrete scale invariance (DSI).","tokens_in":10266,"tokens_out":5175,"duration_ms":50440,"significance":"If the quantitative claims are established, this would be an interesting and novel observation: temporal discrete scale invariance emerging in a homogeneous deterministic lattice model without quenched disorder or an imposed hierarchy, together with a coarsening exponent systematically below the Allen–Cahn value. The qualitative phenomenology — pinned interfaces, abrupt cascades, log-periodic density — is plausible and visually supported. The paper also benefits from large-scale simulations and careful tabulation of fit parameters. However, as argued below, the central DSI interpretation currently rests on an unvalidated exponential-collapse law and on fits whose uncertainties are not reported, so the quantitative conclusions are not yet load-bearing.","major_comments":[{"comment":"The exponential collapse law dL/dt = -v0 e^{-κL} is the load-bearing premise for the entire DSI interpretation, but the Appendix explicitly states that the linearization is 'only qualitative' and that κ 'does not provide a quantitative prediction of cascade times.' No direct measurement of dL/dt versus L, and no direct measurement of successive cascade times t_i, is reported. The quoted λ_t≈3.476 in Eq. (9) is therefore not an independently measured cascade-time ratio; it is inferred from the assumed exponential law combined with the observed L_i growth. To support DSI, the paper should (i) test Eq. (7) by direct measurement of collapsing domain lengths, and (ii) report the measured sequence of cascade times and their ratios, with uncertainties, and compare directly with λ=exp(2π/ω).","section":"§5, Eqs. (7)–(9), Appendix A"},{"comment":"The central quantitative claim δ<0.5 rests on nonlinear least-squares fits in which δ and D_c are free parameters, yet the table gives uncertainties only for ω. No error bars are reported for δ, and no stability analysis is provided (e.g., variation of fit window or initial guesses). In addition, the D_c values are quoted to ten significant digits with no criterion describing how the critical coupling is determined; Section 7 defines D_c only qualitatively as separating pinned and mobile fronts. Without a protocol for D_c or error estimates for δ, the claim of a systematic deviation from δ=1/2 is not quantitatively supported.","section":"Table 1, §4"},{"comment":"The statement that 'the product δω is approximately 2 for all values of α studied' is not supported by the table. From the listed values, δω equals 1.97, 2.38, 2.38, and 1.70 for α=20, 10, 5, and 3 respectively — a spread of roughly 40%. Since δ has no reported uncertainty, this does not reliably suggest an inverse scaling δ∼1/ω. This is a secondary point, but it should be corrected or removed.","section":"§4, Table 1"},{"comment":"The simulation parameters are incompletely specified. The model contains γ and m, but the simulation section states only α, n=m=2, N=50000, Δt=0.01, t_max≈2.5×10^6, and the initial conditions; γ is never given. Unless γ=1 is intended and simply omitted, the simulations are not reproducible. The depinning example in §2.1 also uses D=0.091 without stating how this relates to the reported D_c values.","section":"§2, Eqs. (3)–(4)"}],"minor_comments":[{"comment":"The claim that the slope is 'close to one lattice site per cascade' is not quantified. A linear fit with slope and uncertainty should be reported. Also, individual domain lengths are integers, but the mean length can change by less than one lattice site, so lattice discreteness does not by itself force Δ⟨L⟩=1.","section":"Fig. 5 and §5"},{"comment":"The table caption states that φ is fitted and has uncertainties, but φ is not shown in the table and no error bars are given for it. Please include φ (or state that it is omitted for space) and its uncertainty.","section":"Table 1"},{"comment":"For α=3, the amplitude ratio c1/c0 ≈ 0.05 is very small; the claim of 'clear' log-periodic oscillations would be stronger with residual plots or a comparison to a pure power-law model.","section":"Fig. 4"},{"comment":"There are several typographical and grammatical issues (e.g., 'atleast' in §2, 'theoretical framework predicts that can explain' in §4). The manuscript would benefit from careful proofreading.","section":"Throughout"},{"comment":"No data/code availability statement is provided. Sharing simulation code and the fit data would materially help verification of the quantitative claims.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The qualitative cascade phenomenology appears genuine and worth reporting, but the DSI claim is not yet established. The most important missing piece is a direct, quantitative test of the exponential collapse law and a direct measurement of cascade-time ratios. If the authors can supply those, the paper could be suitable for publication; as it stands, the central quantitative conclusions rest on an explicitly heuristic law and fits without error bars."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nRead the Bhoyar–Gade paper. The short version: the qualitative story is good and likely correct, but the quantitative discrete-scale-invariance claim is carried by an exponential collapse law that the paper's own appendix concedes is only heuristic.\n\nWhat is actually new: for a 1D lattice of bistable toggle units, they show that domain walls pin below a critical coupling D_c, then disappear in discrete cascades, with the density decaying as a power law with log-periodic oscillations. They also report δ < 0.5 for all promoter strengths, and find that the average size of disappearing domains grows roughly linearly with cascade index. That combination—temporal DSI in a clean, disorder-free, hierarchy-free deterministic lattice—is not in the cited literature. The paper is honest about what is established and what is not, and the qualitative phenomenology in the figures looks convincing.\n\nThe soft spots are concentrated in the quantitative claims. The ratio λ_t ≈ 3.476 is not measured from actual cascade times; it is computed by combining the assumed t_c ∼ e^{κL} law with the measured ~1-site-per-cascade growth of disappearing domains. The appendix explicitly says the linearization giving κ is “only qualitative” and “does not provide a quantitative prediction of cascade times.” So the apparent consistency between λ_t and the log-periodic fit λ = e^{2π/ω} is a check between two observables of the same runs, not an independent measurement. Second, the claimed δω ≈ 2 relation is not satisfied by Table 1: for α=10 and 5 the product is ≈2.38, and for α=3 it is ≈1.70. That weakens the inverse-scaling suggestion. Third, there are no error bars on δ or D_c, and D_c is selected as “close to depinning” without a stated criterion. Finally, no code or data are provided, which for a numerical paper of this type is a real impediment to checking the fits.\n\nNone of this kills the core qualitative mechanism. Pinned fronts, abrupt cascade annihilation, and log-periodic density decay are plausible and appear in the spatiotemporal data. The paper deserves a serious referee, but it needs major revision: direct measurement of the cascade time sequence, a principled D_c determination, ensemble error bars, and either a quantitative derivation of the collapse law or a direct numerical test of it. I would not cite it as it stands, but I would take a revised version seriously.\n\nRecommendation: send it to peer review rather than desk rejecting; expect major revision.\n\nBest,\n[You]","headline":"The qualitative cascade picture is probably right, but the quantitative DSI claim is not yet supported: lambda_t is inferred from an unvalidated collapse law, not measured directly.","tokens_in":10800,"tokens_out":2157,"would_cite":false,"duration_ms":22187,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A chain of diffusively coupled bistable gene switches coarsens by discrete cascade collapse, not smooth front motion, with log-periodic oscillations and a slower-than-classical power-law exponent.","keywords":["bistable gene toggle","coarsening","domain walls","cascade dynamics","log-periodic oscillations","discrete scale invariance","front pinning","reaction-diffusion lattice"],"falsifier":"Simulate isolated domains of fixed length L in the same bistable model and measure collapse time as a function of L; if t_c(L) is not exponential in L (or does not match e^{κL}/(κv0) with independently measured κ), the mechanism is wrong. Also test whether t_{i+1}/t_i remains constant over many more cascades than simulated; a visible drift in the ratio would mean the discrete scale invariance is only approximate.","tokens_in":9722,"feed_emoji":"🧬","tokens_out":5279,"duration_ms":46725,"temperature":0.7,"pith_summary":"This paper tries to establish that a one-dimensional lattice of diffusively coupled bistable gene switches coarsens through deterministic cascade events: domain walls stay pinned for long stretches, then whole domains vanish abruptly. It claims the domain-wall density decays as a power law, but with an exponent below the classical value of 1/2 for curvature-driven coarsening, and with log-periodic oscillations superimposed. The oscillations are controlled by the sizes of disappearing domains, which grow by roughly one lattice site per cascade, making successive cascade times geometrically spaced and pointing to temporal discrete scale invariance. If true, the result means coordinated, intermittent reorganization of gene-expression patterns can arise in a homogeneous system without quenched disorder or hierarchical spatial structure.","feed_headline":"Cascade collapse slows coarsening below the standard 1/2 exponent","feed_subtitle":"Domain walls pin, then whole domains vanish in bursts, giving log-periodic oscillations and an unusually slow power-law decay.","key_machinery":"The load-bearing mechanism is the exponential front-interaction law dL/dt = -v0 exp(-κL), obtained heuristically in the appendix by linearizing the reaction term around the low stable state and estimating the interaction of two fronts whose tails decay as exp(-κ|x|), with κ = sqrt(-f'(A_L)/D). This law converts integer-valued, roughly linearly growing disappearing-domain lengths into geometric cascade time ratios t_{i+1}/t_i ~ e^κ. The companion ingredient is the empirical observation that the average disappearing-domain length grows by about one lattice site per cascade, which together with the exponential law yields the constant spacing λ_t and the log-periodic oscillations.","core_discovery":"The central claim is that a deterministically evolving lattice of diffusively coupled bistable gene switches, tuned near a critical diffusion strength, coarsens through a cascade mechanism rather than by continuous curvature-driven interface motion. Domain walls remain pinned for long intervals, then entire domains collapse abruptly. The wall density decays as ρ(t) ~ t^{-δ} with δ systematically below 1/2 for all promoter strengths tested (δ ~ 0.28 for the weakest), and the decay is modulated by log-periodic oscillations. The paper identifies the population of disappearing domains—not the wall density—as the controlling variable: their mean size grows roughly linearly with cascade index, app","pith_inferences":["A sharper test would measure the front-localization parameter κ directly from static front profiles and compare it with the fitted λ_t; the paper's own appendix cautions that its κ estimate is only qualitative, so this is the natural next check.","If the one-site-per-cascade growth persists indefinitely, cascade coarsening maps to an iterated rule L → L+1 with geometric waiting times; that reduction could yield the full distribution of cascade intervals and possibly derive δ from κ and boundary conditions, which the paper leaves open.","The δω ~ 2 product hints at a scaling relation near the depinning threshold that may survive under unequal diffusivities or in higher dimensions; testing whether the cascade picture persists in two dimensions would delineate whether this is a one-dimensional lattice effect or a generic mechanism."],"forward_implications":["If correct, deterministic bistable systems can violate the standard 1/2 coarsening exponent even with no noise, no conserved order parameter, and no curvature-driven dynamics; slow cascade coarsening with δ ~ 0.28 is a candidate for a distinct universality class.","The measured near-constant product δω ~ 2 links the coarsening exponent to the log-periodic frequency, giving a quantitative relation that can be checked in other systems.","Because the mechanism only needs bistability, diffusive coupling, front pinning, and discrete domains, the paper predicts that similar cascade coarsening and temporal discrete scale invariance should appear in other reaction-diffusion and bistable lattice models.","The log-periodicity is tied to the disappearing-domain length spectrum, so observations should focus on collapse events themselves rather than only on interface counts."],"fun_headline_variants":["Cascade collapses slow coarsening, exponent below 1/2","Log-periodic coarsening from abrupt domain cascades","Gene toggle coarsens by cascades: exponent < 0.5","Curvature-free coarsening: cascades, not smooth walls","Abrupt domain death gives log-periodic coarsening"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The argument rests on the assumption that a domain's collapse time grows exponentially with its length (because front interactions decay exponentially) and that the typical disappearing domain grows by about one lattice site per cascade; if either fails outside the fitted regime, the geometric cascade spacing and discrete-scale-invariance picture collapse.","fun_headline_variants_meta":{"raw":{"variants":["Cascade collapses slow coarsening, exponent below 1/2","Log-periodic coarsening from abrupt domain cascades","Gene toggle coarsens by cascades: exponent < 0.5","Curvature-free coarsening: cascades, not smooth walls","Abrupt domain death gives log-periodic coarsening"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000951,"raw_usage":{"total_tokens":3877,"prompt_tokens":710,"completion_tokens":3167,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":454,"completion_tokens_details":{"reasoning_tokens":3076}},"tokens_in":454,"tokens_out":3167,"duration_ms":20103,"temperature":1.0,"reasoning_tokens":3076,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T14:00:48.607860+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate isolated domains of fixed length L in the same bistable model and measure collapse time as a function of L; if t_c(L) is not exponential in L (or does not match e^{κL}/(κv0) with independently measured κ), the mechanism is wrong. Also test whether t_{i+1}/t_i remains constant over many more cascades than simulated; a visible drift in the ratio would mean the discrete scale invariance is only approximate.","supporting_citations":[],"review_version":1}