{"ID":2866189,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.21557","arxiv_id":"2509.21557","title":"Generation-Time vs. Post-hoc Citation: A Holistic Evaluation of LLM Attribution","abstract":"Trustworthy Large Language Models (LLMs) must cite human-verifiable sources in high-stakes domains such as healthcare, law, academia, and finance, where even small errors can have severe consequences. Practitioners and researchers face a choice: let models generate citations during decoding, or let models draft answers first and then attach appropriate citations. To clarify this choice, we introduce two paradigms: Generation-Time Citation (G-Cite), which produces the answer and citations in one pass, and Post-hoc Citation (P-Cite), which adds or verifies citations after drafting. We conduct a comprehensive evaluation from zero-shot to advanced retrieval-augmented methods across four popular attribution datasets and provide evidence-based recommendations that weigh trade-offs across use cases. Our results show a consistent trade-off between coverage and citation correctness, with retrieval as the main driver of attribution quality in both paradigms. P-Cite methods achieve high coverage with competitive correctness and moderate latency, whereas G-Cite methods prioritize precision at the cost of coverage and speed. We recommend a retrieval-centric, P-Cite-first approach for high-stakes applications, reserving G-Cite for precision-critical settings such as strict claim verification. Our codes and human evaluation results are available at https://anonymous.4open.science/r/Citation_Paradigms-BBB5/","short_abstract":"Trustworthy Large Language Models (LLMs) must cite human-verifiable sources in high-stakes domains such as healthcare, law, academia, and finance, where even small errors can have severe consequences. Practitioners and researchers face a choice: let models generate citations during decoding, or let models draft answers...","url_abs":"https://arxiv.org/abs/2509.21557","url_pdf":"https://arxiv.org/pdf/2509.21557v2","authors":"[\"Yash Saxena\",\"Raviteja Bommireddy\",\"Ankur Padia\",\"Manas Gaur\"]","published":"2025-09-25T20:39:26Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[\"Large Language Model\",\"Language Model\"]","project_urls":"[\"https://anonymous.4open.science/r/Citation_Paradigms-BBB5/\"]","has_code":false}