USR-2026-0008 · Dark energy

An oscillating dark energy ties the standard fit

On this data combination the CPL best fit is not uniquely selected: a five-parameter oscillating form ties it (χ² = 1413.72 versus 1414.11). But at its best fit the amplitude is pinned against the code's own A < 1.5 bound with a 24-e-fold period against 1.25 e-folds of data, so the 'oscillation' is a monotone segment — and CPL still wins on AIC by 1.6.

PROVISIONAL χ² tie (1413.72 vs 1414.11) but the oscillator is pinned at a code bound; CPL wins on AIC by 1.6 re-run & confirmed 2026-07-09 (R0 — David Thompson's CI; independence pending)

CORRECTED 2026-07-26: the tie reproduces, but it shows that CPL-like ramps are degenerate with each other, NOT that the data cannot distinguish functional forms. The oscillating fit is a boundary runaway, not a competitive alternative.

Falsify-box — how to kill this claim

UNTESTED Three branches, any one fires. (a) THE TIE ITSELF (the claim's first assertion, which the draft never tested): the best oscillating fit and the CPL best fit cease to tie, |chi^2_osc - chi^2_CPL| > 2, on a refit of run_vibration.py through the same DESI DR2 + Pantheon+ + BBN + CMB-R combination. Current post-Lyman-alpha-fix values: osc 1413.7 versus CPL 1414.1, |Delta chi^2| = 0.4, so this branch is SURVIVED; the registered 1414.1 versus 1414.7 no longer reproduces and must be updated. (b) THE TIE IS A REPARAMETERISATION, NOT A FORM-DEGENERACY -- this has ALREADY FIRED and the row must be corrected before it can be statused. It fires if the oscillating fit's own best parameters show it is not oscillating across the observed range: amplitude A pinned at the code's prior wall (run_vibration.py line 44 rejects A >= 1.5), or A consistent with zero, or period 2*pi/k exceeding the observed baseline ln(1 + z_max) = 1.253 e-folds so less than one cycle is sampled. Verified fired by re-running the claim's own script: A = 1.50 at the wall, k = 0.26, period 24.1 e-folds (5.2% of one cycle), and w(z) monotone across the data, -0.86 at z = 0 through -1.02, -1.13, -1.29 at z = 2 -- a long-period cosine segment, i.e. a re-parameterised CPL-like ramp. (c) THE BROADER GLOSS ('the data cannot distinguish the functional form') dies if any evolving-DE form with at most 4 parameters beats CPL by Delta chi^2 >= 4 on this same combination -- e.g. a transition/bump form, or the binned-w reconstruction behind USR-2026-0004.

Verification record — every quoted number, re-run 2026-07-09

quantity	measured	verdict
oscillating vs CPL χ²	1413.72 vs 1414.11 — a tie	MATCH
oscillator amplitude at best fit	pinned against the code's A < 1.5 bound	MATCH
oscillator period	24 e-folds against 1.25 e-folds of data — a monotone segment	MATCH
AIC verdict	CPL preferred by 1.6	MATCH

Provenance

script: expansion_seam/run_vibration.py
script status: PRESENT
datasets: DESI DR2 BAO · Pantheon+ · Planck CMB (compressed) · claim-scripts (downloadable from /data — run it yourself)
re-run: 2026-07-09

Status, honestly

CORRECTED 2026-07-26 by the kill-test sweep; the audit flag it carried is resolved. This status was assigned by hand from the public re-run audit above — it is a judgement made from that evidence, not a number this site computed for itself. What cannot move it is authority: no vote, endorsement, or say-so, from anyone including the maintainer. Independent reproduction would move it; nothing else will.

§ Reproduction run_vibration.py

David Thompson's own run of this script, captured verbatim. A match proves the result is reproducible; it is still R0 on this registry's independence rings — same code, so it cannot move a status. Only an outside run does that.

sha256 0c7371dfe575a6bb3e0aea0c… size 4179 B exit 0 runtime 87.3s env python 3.12.2 · numpy 1.26.4

A · Recorded David Thompson's machine · 2026-08-02

VIBRATION FIT (w = -1 + A cos(k ln a + phi)):
  LambdaCDM  chi2=1421.1 (2p, AIC=1421.1)
  crossing   chi2=1414.1 (4p, AIC=1418.1)
  vibration  chi2=1413.7 (5p, AIC=1419.7)
  best vibration: amplitude A=1.50, frequency k=0.26 /e-fold,
  period = 24.1 e-folds  ~ 347 billion years;  w(today)=-0.86
    w(z=0.0) = -0.86
    w(z=0.5) = -1.02
    w(z=1.0) = -1.13
    w(z=2.0) = -1.29

B · Yours not runnable in-browser

This script needs data files, a heavy dependency,
or more time than a browser tab should take.
Download it from /data and run it locally.

Share this claim X Bluesky LinkedIn Reddit HN Email