Which is a typical use of BART?

• A. Image classification
• B. Text summarization
• C. Audio transcription
• D. Speech synthesis

Answer: (B)

BART is a text-to-text transformer trained with a denoising objective, so it learns to transform input text into fluent, coherent output. This makes text summarization a natural fit: given longer writing, the model can generate a shorter, information-packed version that preserves the main ideas. The strength comes from its ability to understand context and produce concise rewrites that stay faithful to the original meaning. Other tasks involving different modalities—like processing images for classification, converting spoken language to text, or generating speech from text—rely on models specialized for those inputs or for audio, so they aren’t typical uses for a text-focused model like BART.