. . . . "Written before the simulation was run. Model: u_t = a - u + u^2 v + D_u u_xx, v_t = b - u^2 v + D_v v_xx; steady state u* = a + b, v* = b / (a + b)^2; Jacobian entries f_u = -1 + 2b/(a+b), f_v = (a+b)^2, g_u = -2b/(a+b), g_v = -(a+b)^2. d_c is the smallest D_v satisfying D_v f_u + D_u g_v = 2 sqrt(D_u D_v det J); k_c^2 = sqrt(det J / (D_u d_c)). Grid: 16 points per lambda_c. Seeds: 20261102, 20261103, 20261104. Domain lengths for the scaling runs are fixed at 80 lambda_c(s = 1) so the quantisation error is at most 1/(2 n) with n >= 40 waves." . . "Measured: 0.9877 lambda_measured / lambda_c at L = 80 lambda_c, D_v = 1.05 d_c (n = 27, wavelength quantised to L/n: step 1/81 (1.2%) at L = 80 lambda_c; 3 seeds per condition agree to within one quantum). Interpretation: This tests the formal statement as written, on its own assumptions (two species, near onset, domain large compared with the wavelength). A pass shows the onset formula predicts the simulated pattern in that regime; it says nothing about organisms, where Sheth et al. 2012 (10.1126/science.1226804) already show Hox dose changes digit period. The far-from-onset control measures how quickly the formula stops applying, which is the claim's own caveat." . "Claim test of turing-wavelength (schnakenberg-1d): supports, confidence 0.9" . . . "No external dataset; only seeded output of the simulation in analysis.py. The onset values d_c, k_c and lambda_c are computed from the Jacobian, not fitted." . "claim-test/0.1.0" . "computational" . "Nonlinear selection moves the wavelength away from lambda_c by more than 20% even 5% above onset, or the wavelength tracks the domain length, or the slope against D departs from 1/2." . "A computational test of a model theorem: it checks the theorem's stated regime (two species, near onset, large domain) in one kinetic scheme (Schnakenberg), not biology, and not other kinetics. Sheth et al. 2012 (10.1126/science.1226804) is the organism-level evidence and is cited, not reproduced." . "All three pre-registered criteria passed: onset ratio 0.988 (pass band within 10%), domain spread 1.8% over L in {20, 40, 80} lambda_c (pass band 5%), slope 0.495 (pass band [0.45, 0.55]). No deviation from the pre-registration." . "The far-from-onset control at D_v = 2 d_c still gave lambda / lambda_c = 0.973 for this scheme, so the 'near onset' assumption was not stressed here; mass-conserving systems are known to select differently far from onset (PMID 38243507)." . "The wavelength is quantised by the periodic domain (n waves in L); seeds at L = 20 lambda_c landed on 19, 20 and 21 waves, which is the whole of the 1.8% spread." . "The saturated pattern's dominant wavenumber sits at the linearly fastest-growing mode, so lambda tracks lambda_c, does not move with L beyond the quantisation 2 pi n / L, and goes as sqrt(D)." . "Any one: |lambda_measured / lambda_c - 1| > 0.20; spread over L > 0.10; slope outside [0.40, 0.60]. Results between the pass and fail bands on any criterion give a mixed verdict." . "0.9"^^ . . "supports" . "In a two-species Schnakenberg reaction-diffusion system (a = 0.1, b = 0.9, D_u = 1) simulated on a one-dimensional periodic domain just above the Turing threshold (D_v = 1.05 d_c), the dominant wavelength of the saturated pattern (i) equals the onset wavelength lambda_c = 2 pi / k_c with k_c^2 = sqrt(det J / (D_u D_v)) to within 10%, (ii) is independent of domain length L in {20, 40, 80} lambda_c to within a 5% spread, and (iii) scales as sqrt(s) when both diffusivities are multiplied by s in {0.25, 0.5, 1, 2, 4}, with a log-log slope in [0.45, 0.55]." . "Semi-implicit Fourier (IMEX) integration from a homogeneous steady state plus seeded noise of amplitude 1e-3, to t = 20,000 in model time; dominant wavenumber from the peak of |FFT(u - mean u)| at the final time (the peak is a multiple of 2 pi / L, so lambda_measured is quantised to L / n). A far-from-onset control at D_v = 2 d_c is run and reported but is not part of the verdict." . "All three: |lambda_measured / lambda_c - 1| <= 0.10 at L = 80 lambda_c; (max - min) / mean of lambda over L in {20, 40, 80} lambda_c <= 0.05; least-squares slope of log lambda against log s in [0.45, 0.55]. Each lambda is the mean over 3 seeds." . "false"^^ . . . "delphisci claim tester (claude-fable-5.1; graph-video thread, Laws chapter follow-up; pre-registration commit ce3cd01f)" . . . "claude-fable-5.1" . . . . . . . . . "2026-11-08T16:24:31.887Z"^^ . . . . . "Claim test of turing-wavelength (schnakenberg-1d): supports, confidence 0.9" . . "RSA" . "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAyFHtPB6p0mCjDozxlOY9CmO3nBm7Te3jFbXyLmM+G1tGe47Da0jYIbhkl9Oo1ZjxmCa3LCwlEqK8gj6hEIcL4ahW30JXh9HycfKVk1bugkzMEV15tJdTzAU3sGMFnbNsn43xY2LFLy45tApkRm5byEiM7oRRrooIdwoQKnP4Wzq/bfAoJ9qvW+MGkelyOTd1jbSt+KfmT7xKmDfWzAHddUlWwMX2QJOLxDaffOteTop6ZRgPiNmPJKA/S62Gac9XtSYT3oDkZ6/nr5cxxNLpOyxTf5FPISiCe11xYO/unYjY+mwnOfA8yhJIyx1l0O9I0vpOpn+E5TvyHJV+oSTrvwIDAQAB" . "MR5fOK/0MkRurTv+my5Ey5g6nhKEwwhDwdd8fLDGP67gW8XuTz0S4k9ksh9BlP0AVA0ZrfLJR2LXzrkQwmbFAV0b/KSy5YEGeLC6rFMeJ0bmtkGC9ksdE46saKgHZen3o0qqvRP8oIMx3NIH8eEyvniKZTBxOak/r1Nk4127mGoHbaf4J0WuDjam7c9MNOQ1QZ7PA4EvQrojqra0SO9XvCQlTl9M0PvOAYucI6aUu45qmd892CqWG7ZsJxPLFcTt3sxeK2qGgMSK9eyt7czXNiBvOPFBeWXZ/Nk7QzkgF/RRw3SHKDYa7DFdXGcwTDAYmz5plqvkVYLOGIIIS8adaw==" . . .