Question 15
Domain 2: Fundamentals of Generative AIWhich option is a characteristic of transformer-based language models?
Correct answer: C
Explanation
Transformer-based language models rely on self-attention, which lets each token weigh other tokens in the sequence and capture contextual relationships. This matches the defining mechanism of transformers, unlike older sequence models that process text mainly in order.
Why each option is right or wrong
A. Transformer-based language models use convolutional layers to apply filters across an input to capture local patterns through filtered views.
Convolutional layers apply filters over local regions; transformers are built around self-attention.
B. Transformer-based language models can process only text data.
Transformer models are not limited to text; some foundation models support multimodal capabilities.
C. Transformer-based language models use self-attention mechanisms to capture contextual relationships.
Transformer architectures are defined by the self-attention layer, which computes token-to-token relevance across the full input sequence rather than relying on a strictly sequential recurrence. In the exam context, that mechanism is what lets the model represent context-dependent meaning for each word based on surrounding words, which is the core distinguishing feature of transformer-based language models.
D. Transformer-based language models process data sequences one element at a time in cyclic iterations.
Element-by-element cyclic processing describes recurrent sequence models, not transformers.