· Domain · Claim-safe

SpO2-flat regimes: mixed results, published limits

Restates locked Evidence / workbench numbers only. Not universal performance.

Zero Overshoot does not sell “works on all biomedical streams.” Excellence hunt v2 family results for VitalDB SpO2:

Metric familyResultPublic posture
Oscillation (SpO2)3/6 wins; mean adv −4.3%Not locked as a win
Peak-dev (SpO2)3/6; mixed / negative meanDo not claim universal peak win
Contrast: VitalDB HR peaks7/7; ~+30.6% peak-devStronger domain cue
Contrast: actuator mocks osc30/30; ~+32%Primary industrial story

Why flat regimes hurt the thermostat story

Soft thermostats shine when there is mid-band energy around a normal. Flat SpO2-like streams often lack that structure; time metrics can favor trackers. Spectral calming may still appear in some packs—do not upgrade that to a product claim without your own validation.

Verdict

Primary public story remains industrial / actuator-class mid-band + peaks. SpO2 is a domain nuance on the homepage for a reason. Validate on your trace class.

Sources: excellence hunt v2 family tables · EXCELLENCE_FINDING_20260804 · Evidence domain notes

Questions people actually ask

Why mention SpO2-flat regimes?

As an honest limit: adaptive x* and aggressive spring behavior can hurt flat traces. Not a medical device claim.

Is GSRF a medical SpO2 filter?

No medical device claims. Flat-regime probes are characterization limits, not clinical validation.

Should I use adaptive x* on flat signals?

Often no — locked caution says adaptive can hurt SpO2-like flat series.

Related research