![[2511.15719] Chain of Summaries: Summarization Through Iterative Questioning](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2511.15719] Chain of Summaries: Summarization Through Iterative Questioning
Summary
The paper presents 'Chain of Summaries' (CoS), a novel method for generating concise, information-dense summaries from web content, enhancing accessibility for Large Language Models (LLMs).
Why It Matters
As LLMs increasingly rely on external web content, the need for digestible formats becomes critical. CoS addresses this by refining summaries through iterative questioning, improving comprehension and usability for both AI and human users. This advancement could significantly enhance how information is processed and utilized in AI applications.
Key Takeaways
- CoS improves summarization by iteratively refining content through questioning.
- The method outperforms existing summarization techniques by significant margins.
- CoS-generated summaries are more efficient, requiring fewer tokens while enhancing Q&A performance.
- The approach is agnostic to specific LLMs, making it versatile for various applications.
- Website maintainers can use CoS to make content more accessible for LLMs.
Computer Science > Artificial Intelligence arXiv:2511.15719 (cs) [Submitted on 12 Nov 2025 (v1), last revised 16 Feb 2026 (this version, v2)] Title:Chain of Summaries: Summarization Through Iterative Questioning Authors:William Brach, Kristián Košťál, Lukas Galke Poech View a PDF of the paper titled Chain of Summaries: Summarization Through Iterative Questioning, by William Brach and 2 other authors View PDF HTML (experimental) Abstract:Large Language Models (LLMs) are increasingly using external web content. However, much of this content is not easily digestible by LLMs due to LLM-unfriendly formats and limitations of context length. To address this issue, we propose a method for generating general-purpose, information-dense summaries that act as plain-text repositories of web content. Inspired by Hegel's dialectical method, our approach, denoted as Chain of Summaries (CoS), iteratively refines an initial summary (thesis) by identifying its limitations through questioning (antithesis), leading to a general-purpose summary (synthesis) that can satisfy current and anticipate future information needs. Experiments on the TriviaQA, TruthfulQA, and SQUAD datasets demonstrate that CoS outperforms zero-shot LLM baselines by up to 66\% and specialized summarization methods such as Chain of Density, BRIO and PEGASUS by up to 27\%. CoS-generated summaries yield higher Q\&A performance compared to the source content, while requiring substantially fewer tokens and being agnostic to th...
