If you believe PES fails under a specific condition, tell us. We will test it.

The Presence Entropy Score separates human motion from synthetic motion with Cohen's d = 2.1 and AUC = 0.94. But every benchmark has boundary conditions it has not yet encountered. If you can describe a scenario where you believe the PES would misclassify — a specific population, a specific motion pattern, a specific hardware configuration — we will design an experiment, run it, and publish the result whether it confirms or refutes the current benchmark.