Astra Scientist

A VIRTUAL COLD-ATOM LAB

Ten measurements.
What would you test next?

Does a temperature difference simply relax—or need a sustained heating source?

Preparing the live lab… Saved cases can be explored immediately.
New observations stay in this browser.
Download a copy to keep them elsewhere.

SAVED REPLAY

Loading the first observation…

measurements shown
Observed cloud

Display shows up to 2,000 saved atoms; all atoms enter the statistics.

Measured temperature differenceError bars: one standard error

Measurement notebook

Open a measurement to inspect its settings and evidence.

Recorded conclusion

Take the evidence with you

Public experiment rules and limitations

Each measurement prepares a fresh cloud. The ten-measurement limit includes the preset initial observation. Both candidate mechanisms permit common heating and an initial temperature imbalance. A falling difference alone does not identify its cause.

Compression c changes the starting anisotropy and relaxation scale. Atom number q changes both relaxation and detected count. Keep these control groups separate; the errors assume independent Gaussian samples and known calibration. “Insufficient evidence” is a valid conclusion.

The four saved cases were selected from development records, not sampled for a performance estimate. Cases A–C contain post-run numerical reviews; Case D is a short interface QA record. The failure case is retained.

Refreshing preserves saved observations for read-only replay; it does not resume an interrupted acquisition. Browser storage can be cleared by your browser—download the original observations for a portable copy.

Astra Scientist in 90 seconds

Edited real application screenshots with locally synthesized narration.