Adaptive Retrieval for Reasoning-Intensive Retrieval
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913028505600000 |
|---|---|
| author | Kim, Jongho Kim, Jaeyoung Hwang, Seung-won Kim, Jihyuk Kim, Yu Jin Lee, Moontae |
| author_facet | Kim, Jongho Kim, Jaeyoung Hwang, Seung-won Kim, Jihyuk Kim, Yu Jin Lee, Moontae |
| contents | We study leveraging adaptive retrieval to ensure sufficient "bridge" documents are retrieved for reasoning-intensive retrieval. Bridge documents are those that contribute to the reasoning process yet are not directly relevant to the initial query. While existing reasoning-based reranker pipelines attempt to surface these documents in ranking, they suffer from bounded recall. Naive solution with adaptive retrieval into these pipelines often leads to planning error propagation. To address this, we propose REPAIR, a framework that bridges this gap by repurposing reasoning plans as dense feedback signals for adaptive retrieval. Our key distinction is enabling mid-course correction during reranking through selective adaptive retrieval, retrieving documents that support the pivotal plan. Experimental results on reasoning-intensive retrieval and complex QA tasks demonstrate that our method outperforms existing baselines by 5.6%pt. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_04618 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Adaptive Retrieval for Reasoning-Intensive Retrieval Kim, Jongho Kim, Jaeyoung Hwang, Seung-won Kim, Jihyuk Kim, Yu Jin Lee, Moontae Information Retrieval We study leveraging adaptive retrieval to ensure sufficient "bridge" documents are retrieved for reasoning-intensive retrieval. Bridge documents are those that contribute to the reasoning process yet are not directly relevant to the initial query. While existing reasoning-based reranker pipelines attempt to surface these documents in ranking, they suffer from bounded recall. Naive solution with adaptive retrieval into these pipelines often leads to planning error propagation. To address this, we propose REPAIR, a framework that bridges this gap by repurposing reasoning plans as dense feedback signals for adaptive retrieval. Our key distinction is enabling mid-course correction during reranking through selective adaptive retrieval, retrieving documents that support the pivotal plan. Experimental results on reasoning-intensive retrieval and complex QA tasks demonstrate that our method outperforms existing baselines by 5.6%pt. |
| title | Adaptive Retrieval for Reasoning-Intensive Retrieval |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2601.04618 |