A neural network architecture that uses self-attention mechanisms to process sequential data efficiently. Transformers revolutionized NLP by enabling parallel processing and capturing long-range dependencies, forming the foundation for modern language models like GPT and BERT systems.
