Frozen result
suppression_min
31%
suppression_max
60%
ky_min
0.1
ky_max
6.0
mu
400
What was computed
Linear CGYRO growth-rate spectrum sweeping kᵧ from the ion scale into the electron band for NT vs PT. NT γ is 31–60% below PT at every resolved kᵧ (0.1–6.0), so the quiet-basin advantage holds across the whole linear spectrum (kᵧ ≥ 10 excluded as unconverged).
Linear γ(kᵧ) spectrum — NT vs PT
| kᵧ | NT γ | PT γ | NT below PT |
|---|---|---|---|
| 0.1 | 0.0369 | 0.0660 | 44% |
| 0.2 | 0.0690 | 0.1210 | 43% |
| 0.3 | 0.0439 | 0.1099 | 60% |
| 0.5 | 0.0702 | 0.1012 | 31% |
| 0.7 | 0.1439 | 0.2250 | 36% |
| 1 | 0.2105 | 0.4209 | 50% |
| 2 | 0.5550 | 0.9642 | 42% |
| 4 | 1.0046 | 1.5423 | 35% |
| 6 | 0.5921 | 1.3909 | 57% |
NT γ is 31–60% below PT at every resolved kᵧ from the ion scale into the low-electron band — the quiet-basin advantage holds across the whole linear spectrum. Honest exclusion: kᵧ ≥ 10 is flagged UNCONVERGED (short integration time t_final, non-physical γ) and is left out of the plot and the claim; only the converged kᵧ = 0.1–6.0 band is used.
Reproduce this
# Reproduce BR-L2-A18c against the frozen record pip install kronos-toolkit # github.com/KronosFE/kronos-toolkit python -c "from kronos_toolkit.verify import run_regression; run_regression()" # BR-L2-A18c is anchored in the 2026-08-27 freeze spine (verdict: reproduced).