The capacitor on the feedback path looks like it is there to reinforce a voltage change that has just made it to the switching threshold on the inverters. If this is true, it would help overcome the limited voltage gain of the inverters.
I think this circuit acts as some kind of restricting amplifier comparator . On output you should have high rise/fall time pulses when opamp output is floating around logic switch threshold voltage .
I think the feedback capacitor is like a speedup capacitor since it will act as a short then the inverter output is transitioning and thus will reinforce the comparator output.
During simulation such circuit behave like filter. Before transition, OTA charges grounded and feedback capacitors (propagation delay increased), when inverters begin act as amplifiers comparator switch faster.
I'd like to make more precise, can such circuit has been named as comparator with noise filter?
It's a positive feedback loop , to speed up the transition with hysteresis.
The first stage is an OTA, charging or discharging the first capacitor, depending on the input voltages. The voltage will rise or fall as the total output capacitance is charged/discharged, until the two cascaded inverters reach the high-gain region and "switch over" , where the output will pull up (or down) the input through the feedback capacitance.