Query-Aware Source-Risk Triage for Retrieval-Augmented Generation

arXiv cs.AI Papers

Summary

The paper proposes a query-aware triage layer for retrieval-augmented generation (RAG) pipelines to handle the material relationship of sources to queries, using scoring and routing methods evaluated on synthetic data to define risk-coverage targets.

arXiv:2609.16564v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) pipelines may omit a source's material relationship to the query. We study a pre-generation triage layer that treats this relationship as query dependent. The method routes canonical query families for enhanced review and assigns retrieved pages to pass, contextualize, exclude, or review. It combines a four-dimension page score, rank-discounted family aggregation, intent-preserving query mutations, and a family-held-out router. A single-coded pilot of 200 real URLs supplies provisional calibration anchors; a 20,000-row scenario with synthetic domain identifiers supports controlled workload analysis. An oracle page gate defines a risk-coverage target for a future learned classifier. The evaluation shows why page-level frequency cannot substitute for family-level exposure and quantifies how calibration changes scenario activation. Annotation reliability remains unmeasured, and synthetic rankings omit real retrieval dynamics. The result is an auditable triage method and validation plan, not an estimate of deployed review workload, live-Web prevalence, or downstream answer-quality gains.
Original Article
View Cached Full Text

Cached at: 09/16/26, 09:00 AM

# Query-Aware Source-Risk Triage for Retrieval-Augmented Generation
Source: [https://arxiv.org/abs/2609.16564](https://arxiv.org/abs/2609.16564)
[View PDF](https://arxiv.org/pdf/2609.16564)

> Abstract:Retrieval\-augmented generation \(RAG\) pipelines may omit a source's material relationship to the query\. We study a pre\-generation triage layer that treats this relationship as query dependent\. The method routes canonical query families for enhanced review and assigns retrieved pages to pass, contextualize, exclude, or review\. It combines a four\-dimension page score, rank\-discounted family aggregation, intent\-preserving query mutations, and a family\-held\-out router\. A single\-coded pilot of 200 real URLs supplies provisional calibration anchors; a 20,000\-row scenario with synthetic domain identifiers supports controlled workload analysis\. An oracle page gate defines a risk\-coverage target for a future learned classifier\. The evaluation shows why page\-level frequency cannot substitute for family\-level exposure and quantifies how calibration changes scenario activation\. Annotation reliability remains unmeasured, and synthetic rankings omit real retrieval dynamics\. The result is an auditable triage method and validation plan, not an estimate of deployed review workload, live\-Web prevalence, or downstream answer\-quality gains\.

## Submission history

From: Lu Yi \[[view email](https://arxiv.org/show-email/e4c89fb0/2609.16564)\] **\[v1\]**Tue, 15 Sep 2026 03:01:27 UTC \(775 KB\)

Similar Articles

Claim-Selective Certification for High-Risk Medical Retrieval-Augmented Generation

arXiv cs.CL

This paper proposes claim-selective certification for high-risk medical retrieval-augmented generation (RAG), decomposing responses into verifiable claims and scoring them against evidence to produce actions (full, partial, conflict, abstain) using an intent-aware selector, achieving low unsupported-claim risk and high action accuracy.