نسخة أولية وصول مفتوح
Rethinking Faithfulness in LLMs: A Pairwise Context-Sensitive Perspective
Large language models (LLMs) are expected to answer questions faithfully based on the provided context, abstaining when the context information is insufficient to answer the questions. Existing faithfulness evaluations typically assess each question-context instance in isolation; however, such instance-level evaluation …