Handling extremely large documents for question answering is challenging: chunk-based embedding methods often lose track of important global context, while full-context transformers can be prohibitively expensive for hundreds of thousands of tokens. We propose a single-pass document scanning approach that processes the entire text in linear time, preserving global coherence while deciding which sentences are most relevant to the query. On 41 QA benchmarks, our single-pass scanner consistently outperforms chunk-based embedding methods and competes with large language models at a fraction of the computational cost. By conditioning on the entire preceding context without chunk breaks, the method preserves global coherence, which is especially important for long documents. Overall, single-pass document scanning offers a simple solution for question answering over massive text. All code, datasets, and model checkpoints are available at https://github.com/MambaRetriever/MambaRetriever
@article{arxiv.2504.03101,
title = {Single-Pass Document Scanning for Question Answering},
author = {Weili Cao and Jianyou Wang and Youze Zheng and Longtian Bao and Qirui Zheng and Taylor Berg-Kirkpatrick and Ramamohan Paturi and Leon Bergen},
journal= {arXiv preprint arXiv:2504.03101},
year = {2025}
}
Comments
Published at Conference on Language Modeling (COLM), 2025