Metraipolis › archive › dc_994fa89cd2a9

ORDER: Task-Conditioned Routing for Retrieval-Augmented Generation — arXiv:2609.17012

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.17012
License (as archived): CC0-1.0
Archived: 2026-10-15T18:32:18.432Z
SHA-256 of the archived text: b075bbb9c2360303ee6f2fe0c55bc0cb4dfe8ef39925a9f516df7d3b38970088
Stable URL of this page: https://metraipolis.com/documento/dc_994fa89cd2a9

How to cite

ORDER: Task-Conditioned Routing for Retrieval-Augmented Generation — arXiv:2609.17012 — arXiv (en), https://arxiv.org/abs/2609.17012 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_994fa89cd2a9 (Metraipolis archive dc_994fa89cd2a9, sha256 b075bbb9c2360303ee6f2fe0c55bc0cb4dfe8ef39925a9f516df7d3b38970088)

BibTeX

@misc{dc994fa89cd2a9,
  title = {ORDER: Task-Conditioned Routing for Retrieval-Augmented Generation — arXiv:2609.17012},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.17012},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_994fa89cd2a9; sha256 b075bbb9c2360303ee6f2fe0c55bc0cb4dfe8ef39925a9f516df7d3b38970088},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.17012. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Retrieval-Augmented Generation (RAG) pipelines typically rely on a fixed indexing and retrieval configuration determined at preprocessing time. This one-size-fits-all design is ill-suited to domain-expert settings, where heterogeneous queries require different chunking granularities, metadata constraints, and source-selection strategies. As a result, configurations that are effective for one family of queries often perform poorly for others. In this paper, we introduce ORDER (Optimal Routing for Dynamic Evidence Retrieval), a query-conditioned RAG framework that jointly adapts indexing and retrieval to the incoming query. Our approach first discovers semantic clusters over a given set of questions associated to a corpus and learns, for each cluster, a chunking strategy together with a suited metadata filtering and reranking configuration. At inference time, queries are routed to the appropriate pre-built index through nearest-centroid assignment. To further improve retrieval, we propose a supervised query router (QRe) that predicts which collections are most likely to contain relevant evidence, coupled with a Uniform Multi-source Sampler (UMS) that allocates the retrieval budget evenly across the selected sources. We evaluate our framework on large-scale, heterogeneous historical archives and show that conditioning both indexing and retrieval on the query consistently outperforms both naive baselines and strong state-of-the-art RAG systems in complex expert-domain environments.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-16 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.17012 — 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 2026-10-15T18:03:09.498Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-16?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).