# QA-prompting: Improving Summarization with Large Language Models using Question-Answering Source: [QA-prompting: Improving Summarization with Large Language Models using Question-Answering]() Collection index: [llms.txt]() Data updated: 2025-10-26T05:41:29.127Z Authors: Neelabh Sinha Published: 2025-05-20T13:29:36.000Z ## Abstract Language Models (LMs) have revolutionized natural language processing, enabling high-quality text generation through prompting and in-context learning. However, models often struggle with long-context summarization due to positional biases, leading to suboptimal extraction of critical information. There are techniques to improve this with fine-tuning, pipelining, or using complex techniques, which have their own challenges. To solve these challenges, we propose QA-prompting - a simple prompting method for summarization that utilizes question-answering as an intermediate step prior to summary generation. Our method extracts key information and enriches the context of text to mitigate positional biases and improve summarization in a single LM call per task without requiring fine-tuning or pipelining. Experiments on multiple datasets belonging to different domains using ten state-of-the-art pre-trained models demonstrate that QA-prompting outperforms baseline and other state-of-the-art methods, achieving up to 29% improvement in ROUGE scores. This provides an effective and scalable solution for summarization and highlights the importance of domain-specific question selection for optimal performance. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Scientific Document Summarization on PubMed](): Scientific Document Summarization; Pubmed; ROUGE-1: 0.2634. - [Scientific Document Summarization on ArXiv (R-1, R-2, R-L, METEOR, BERTScore)](): Scientific Document Summarization; arXiv; ROUGE-1: 26.35. - [Opinion Summarization on Rotten Tomatoes (full dataset)](): Opinion Summarization; Rotten Tomatoes (full dataset); Coverage: 5.3. ## Original sources and implementation [Original paper]() [PDF]()