Back to age-quantum.org AGE-TGI-2S3 + QEC · Photonic quantum research · ES/EN

Uncategorized

Refute AGE: five predictions your lab can test

Refute AGE: five predictions your lab can test
predicciones computacionales falsables

Everything AGE argues today comes from simulation. It is reproducible and every source is cited, but it is still simulation. A scientific claim starts to count when someone who did not write it tries to knock it down. This post invites you to do exactly that.

Five concrete predictions

  1. WDM drift. On a serially monitored bank, joint estimation reduces error compared with per-ring estimation on a time segment not used for fitting. If the real gain stays at per-ring level (~1.06×), our common-mode hypothesis fails. Benchmark.
  2. Realizability. With ρ_B = 1.279, the C2⋆ receiver is executable without clipping the control; C3 needs ρ ≈ 1.86. TM-02.
  3. Offset invariance. C2⋆ capture does not change when a common carrier offset rises from 1.75 to 3.5 MHz. TM-02.
  4. Calibration cost. Selectivity grows in proportion to shots per fit, up to a crosstalk floor, with no absolute phase reference. TM-03.
  5. Control budget. C2⋆ needs peak coupling near 0.5 Γ and a control bandwidth of at least 15 Γ. TM-04.

How we do it without asking for your trust

For photonic chip logs we use the free evaluation protocol: we agree on metric, baseline and failure criterion; fit only on the allowed window; seal the forecast with a SHA-256 hash before seeing the held-out segment; then score. Your data can stay on your machine. For the temporal-mode benches, each simulator’s JSON export serves as a pre-registration of the prediction.

There are three possible outcomes and all three get published: confirmation, inconclusive, or refutation. A well-documented refutation saves us years of work in the wrong direction.

What we offer

Open code with a DOI, support to adapt the analysis to your platform, and shared authorship on the publication of the result, whatever it is. The platforms we are looking for are on Platforms. All current evidence, with its status, is on Evidence.

If you are new here, start with why we measure before integrating.

CTA: Propose a validation — describe the platform and the log; do not send data in the first contact.

// Discussion 0 replies

Leave a reply

Your email will not be published. Fields marked with * are required.