Hard Negative
A training example that is visually or semantically very similar to the anchor but belongs to a different class, forcing the model to learn fine details.
Think of It Like This
Like identical twins where you have to learn subtle differences to tell who is who.
Hard negatives are crucial in contrastive learning to prevent models from taking shortcuts by relying on obvious features. By presenting examples that are difficult to distinguish, the model is forced to learn more robust and discriminative representations. This significantly improves performance on complex boundary cases.