Binary activation functions (BAFs) play as a unique and intriguing class within the realm of machine learning. These activations possess the distinctive property of outputting either a 0 or a 1, representing an on/off state. This simplicity makes them particularly attractive for applications where binary classification is the primary goal. While B