Metraipolis › archive › dc_62012b0ddc31
When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1 — arXiv:2608.13237
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.13237
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T04:24:43.109Z
- SHA-256 of the archived text
5707f127f8e95c095dcd557696d5d8c3012be58cf5f3055e5c758b726fb235c5- Stable URL of this page
https://metraipolis.com/documento/dc_62012b0ddc31
How to cite
When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1 — arXiv:2608.13237 — arXiv (en), https://arxiv.org/abs/2608.13237 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_62012b0ddc31 (Metraipolis archive dc_62012b0ddc31, sha256 5707f127f8e95c095dcd557696d5d8c3012be58cf5f3055e5c758b726fb235c5)
BibTeX
@misc{dc62012b0ddc31,
title = {When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1 — arXiv:2608.13237},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.13237},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_62012b0ddc31; sha256 5707f127f8e95c095dcd557696d5d8c3012be58cf5f3055e5c758b726fb235c5},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.13237. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-round retrieval-augmented generation (RAG) must decide when to stop searching as evidence accumulates. Because the deployed policy is determined by the first STOP on each trajectory, this is a sequential selection problem rather than an independent state-classification task. We adapt S2G-RAG's structured sufficiency-and-gap judgment to a frozen Search-R1 pipeline and train a Qwen3.5-2B judge on 3,009 states from 900 disjoint HotpotQA questions. Search-R1's reasoner, retriever, corpus, prompt, and search budget remain unchanged, while the judge checkpoint and stopping threshold are selected on grouped validation and frozen before confirmatory evaluation. On the confirmatory test set, the resulting policy reduces retrieval calls by 77 (3.70\%) relative to Native Search-R1, while Official Exact Match decreases by 0.625 percentage points. Thus, the trained S2G-style structured judge reduces retrieval while broadly preserving answer accuracy. The result does not imply unchanged or improved accuracy, safe stopping, or lower total inference cost. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.CL). Categorie arXiv: Information Retrieval (cs.IR) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.13237 — 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 2026-10-19T04:17:42.415Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).