Knowing When Not to Answer: Lightweight KB-Aligned OOD Detection for Safe RAG
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912830843781120 |
|---|---|
| author | Triantafyllopoulos, Ilias Qu, Renyi Giorgi, Salvatore Curtis, Brenda Ungar, Lyle H. Sedoc, João |
| author_facet | Triantafyllopoulos, Ilias Qu, Renyi Giorgi, Salvatore Curtis, Brenda Ungar, Lyle H. Sedoc, João |
| contents | Retrieval-Augmented Generation (RAG) systems are increasingly deployed in high-stakes domains, where safety depends not only on how a system answers, but also on whether a query should be answered given a knowledge base (KB). Out-of-domain (OOD) queries can cause dense retrieval to surface weakly related context and lead the generator to produce fluent but unjustified responses. We study lightweight, KB-aligned OOD detection as an always-on gate for RAG systems. Our approach applies PCA to KB embeddings and scores queries in a compact subspace selected either by explained-variance retention (EVR) or by a separability-driven t-test ranking. We evaluate geometric semantic-search rules and lightweight classifiers across 16 domains, including high-stakes COVID-19 and Substance Use KBs, and stress-test robustness using both LLM-generated attacks and an in-the-wild 4chan attack. We find that low-dimensional detectors achieve competitive OOD performance while being faster, cheaper, and more interpretable than prompted LLM-based judges. Finally, human and LLM-based evaluations show that OOD queries primarily degrade the relevance of RAG outputs, showing the need for efficient external OOD detection to maintain safe, in-scope behavior. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_02296 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Knowing When Not to Answer: Lightweight KB-Aligned OOD Detection for Safe RAG Triantafyllopoulos, Ilias Qu, Renyi Giorgi, Salvatore Curtis, Brenda Ungar, Lyle H. Sedoc, João Computation and Language Information Retrieval Retrieval-Augmented Generation (RAG) systems are increasingly deployed in high-stakes domains, where safety depends not only on how a system answers, but also on whether a query should be answered given a knowledge base (KB). Out-of-domain (OOD) queries can cause dense retrieval to surface weakly related context and lead the generator to produce fluent but unjustified responses. We study lightweight, KB-aligned OOD detection as an always-on gate for RAG systems. Our approach applies PCA to KB embeddings and scores queries in a compact subspace selected either by explained-variance retention (EVR) or by a separability-driven t-test ranking. We evaluate geometric semantic-search rules and lightweight classifiers across 16 domains, including high-stakes COVID-19 and Substance Use KBs, and stress-test robustness using both LLM-generated attacks and an in-the-wild 4chan attack. We find that low-dimensional detectors achieve competitive OOD performance while being faster, cheaper, and more interpretable than prompted LLM-based judges. Finally, human and LLM-based evaluations show that OOD queries primarily degrade the relevance of RAG outputs, showing the need for efficient external OOD detection to maintain safe, in-scope behavior. |
| title | Knowing When Not to Answer: Lightweight KB-Aligned OOD Detection for Safe RAG |
| topic | Computation and Language Information Retrieval |
| url | https://arxiv.org/abs/2508.02296 |