# When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1

arXiv: https://arxiv.org/abs/2608.13237 · HTML: https://mlsanity.com/paper/2608.13237/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation.
- Builds on 10 of its 14 references that are in the map.
- First posted on arXiv 13 Aug 2026.

## Placed in

- Retrieval-augmented generation / [RAG robustness and adaptive retrieval](https://mlsanity.com/rag/rag-robustness-and-adaptive-retrieval/)

## Builds on (most cited first, 10 of 10)

- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025)
- [Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions](https://mlsanity.com/paper/2212.10509/) (2022)
- [Selective Classification for Deep Neural Networks](https://mlsanity.com/paper/1705.08500/) (2017)
- [Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity](https://mlsanity.com/paper/2403.14403/) (2024)
- [C-RAG: Certified Generation Risks for Retrieval-Augmented Language Models](https://mlsanity.com/paper/2402.03181/) (2024)
- [Knowing You Don't Know: Learning When to Continue Search in Multi-round RAG through Self-Practicing](https://mlsanity.com/paper/2505.02811/) (2025)
- [S2G-RAG: Structured Sufficiency and Gap Judging for Iterative Retrieval-Augmented QA](https://mlsanity.com/paper/2604.23783/) (2026)
- [TASR: Training-Free Adaptive Stopping for Iterative Retrieval](https://mlsanity.com/paper/2606.13814/) (2026)

## Abstract (arXiv)

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.