Metraipolis › archive › dc_d31ec0c8d2d3
REDSI: Addressing the Reproducibility and Evaluation Consistency of Differentiable Search Indexing for Document Retrieval — arXiv:2609.08860
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.08860
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T15:05:27.093Z
- SHA-256 of the archived text
e3e5e810328cf6d91d96ca888ff40f9c7cc64149a2bae425b929806e535662bb- Stable URL of this page
https://metraipolis.com/documento/dc_d31ec0c8d2d3
How to cite
REDSI: Addressing the Reproducibility and Evaluation Consistency of Differentiable Search Indexing for Document Retrieval — arXiv:2609.08860 — arXiv (en), https://arxiv.org/abs/2609.08860 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d31ec0c8d2d3 (Metraipolis archive dc_d31ec0c8d2d3, sha256 e3e5e810328cf6d91d96ca888ff40f9c7cc64149a2bae425b929806e535662bb)
BibTeX
@misc{dcd31ec0c8d2d3,
title = {REDSI: Addressing the Reproducibility and Evaluation Consistency of Differentiable Search Indexing for Document Retrieval — arXiv:2609.08860},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.08860},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d31ec0c8d2d3; sha256 e3e5e810328cf6d91d96ca888ff40f9c7cc64149a2bae425b929806e535662bb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.08860. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
The differentiable search index (DSI) framework (Tay et al., 2022) has become the de facto baseline for generative retrieval. However, DSI is hard to reproduce: no public implementation covers all three original document identifier types (atomic, naive, semantic), reported results vary widely, and the ubiquitous NQ320K dataset is built from Natural Questions through diverse and underspecified preprocessing. We introduce ReDSI, the first open-source DSI implementation supporting all three identifier types, together with a parameterizable and well-documented NQ320K construction pipeline. Experimentally, we achieve results that are competitive with or stronger than previous DSI baselines. Moreover, we conduct extensive experiments under model downscaling, covering retrieval effectiveness, parameter efficiency, training methods and decoding strategies, opening novel directions for future research. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-09 in cs.IR). Categorie arXiv: Information Retrieval (cs.IR). Pagina del lavoro: https://arxiv.org/abs/2609.08860 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-16T14:49:33.967Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).