question: What is the purpose of the Triplet Loss? option 1: To compare an input to a positive and a negative input and maximize association between positively associated inputs option 2: To enhance multiple loss terms with a regularization term option 3: To ensure diversity of actions in reinforcement learning option 4: To penalize the discriminator in generative adversarial networks option 5: To drop correlated areas of feature maps in convolutional neural networks 