Sentence Reordering as an Input Corruption Method

Document Rotation as an Input Corruption Method

A research team is training a model on multi-paragraph documents. Their primary goal is to ensure the model learns the logical flow and coherence between sentences, not just the relationships between words within a single sentence. Which of the following input corruption strategies is specifically designed to target this higher-level, inter-sentence understanding?

Rationale for Sentence-Level Corruption

A language model is being trained using a denoising objective, where it learns to reconstruct original text from a corrupted version. Match each type of input corruption with the primary linguistic feature it forces the model to learn.

Sentence Reordering as an Input Corruption Method

A pre-training process is applied to the following two-sentence input: 'The team celebrated their victory. They had trained hard for months.' The input is transformed into: 'The team [MASK] hard for months.' The model is then tasked with reconstructing the original, complete text from this transformed input. Which specific data corruption technique, designed for handling sequences of text, does this process exemplify?

Choosing a Pre-training Strategy

A model is pre-trained using corruption techniques that operate on the structure of multi-sentence documents. Match each technique with its correct description.