A superconducting detector usually works like an honest ruler: absorb a photon, get a pulse whose amplitude is strictly proportional to the energy. This mode is convenient, but it has a ceiling: a weak signal easily gets lost in the amplifier noise.

Now imagine a swing. While you pump it gently, it obediently returns to the bottom point. But if you push harder, the swing reaches the top point and hangs there — the system transitions to a new stable state. Exactly this happens to the detector when the readout power becomes large. The current in the resonator changes its inductance, and the response stops being a simple copy of the absorbed pulse.

The authors didn't just describe stationary states; they traced the dynamics of a finite pulse: how it rises, distorts, and decays. It turned out that small pulses have a slower rise, while large ones develop a "shelf" before the fall. Beyond the critical point, a sufficiently energetic pulse switches the detector to another stable branch: it "latches" and stays there after the excitations vanish.

This regime had not been demonstrated experimentally before. The detector transforms from an analog energy meter into a trigger with a tunable threshold. Already today, this can be used in systems limited by amplifier noise: a stretched and amplified pulse is easier to pick out from the noise background. And the latching regime opens the way to single-photon counters with a dynamically tunable threshold — for example, for searching for rare events in astrophysics.

The swing that used to just rock back and forth can now be tuned so that it flips only from a strong push. And that very flip becomes the signal.