Coreference Resolution and Text Entailment
摘要
This chapter examines the development of systems for coreference resolution and text entailment, utilizing LLMs such as BART, RoBERTa, and DeBERTa, as well as SLMs, like DistilRoBERTa, TinyBERT, and MiniLM. These tasks are critical for deep language understanding—coreference resolution identifies relationships between entities across a text, while text entailment determines whether one sentence logically follows from another.