Low-frequency output fluctuations in an open exclusion process with particle pausing

What the math gives to ML

The paper identifies a finite-size noise crossover caused by slow reversible internal states: when the mean number of paused particles is of order one, rare pauses alternate with pause-free transport and large jammed clusters. Its strongest transferable mechanism is the scaling variable Np = L kp / ku, rather than the microscopic pause rate alone, together with a nonmonotonic low-frequency Fano factor and a correlation time controlled by slow-state residence times. A neural analogue is a recurrent or state-space layer with reversible active and paused units, tuned so that the expected number of paused units is near the order-one crossover; this predicts a measurable transition in long-horizon output noise and memory.