Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914608661397504 |
|---|---|
| author | O, Sunah Lee, Jay-Yoon |
| author_facet | O, Sunah Lee, Jay-Yoon |
| contents | In multimodal multi-hop question answering, we focus on the initial retrieval stage via two distinct tasks: (1) evidence set completion, retrieving missing evidence given context, and (2) sequential pool construction, iteratively building the top-$K$ pool from the scratch. Under these settings, we point out that conventional iterative retrieval frameworks often suffer from Semantic Anchoring, where previously fetched evidence traps the retriever and yields entity-centric redundancy. To break this trap, we propose GRAIL (Gap-aware Retrieval via Adaptive Implicit Localization), a paradigm that performs implicit query rewriting directly at the embedding level. By context-subtractive query steering, GRAIL excels at compositional cross-modal reasoning, while additive embedding updates show strength on localized information aggregation. By dynamically routing queries based on task type, our Hybrid Framework achieves a 40.3\% macro-averaged performance gain on MultimodalQA. Extensive evaluations demonstrate that sequential GRAIL retrieves in a superior, noise-resilient manner, significantly expanding the search horizon through iterative gap-aware optimization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_28641 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA O, Sunah Lee, Jay-Yoon Information Retrieval In multimodal multi-hop question answering, we focus on the initial retrieval stage via two distinct tasks: (1) evidence set completion, retrieving missing evidence given context, and (2) sequential pool construction, iteratively building the top-$K$ pool from the scratch. Under these settings, we point out that conventional iterative retrieval frameworks often suffer from Semantic Anchoring, where previously fetched evidence traps the retriever and yields entity-centric redundancy. To break this trap, we propose GRAIL (Gap-aware Retrieval via Adaptive Implicit Localization), a paradigm that performs implicit query rewriting directly at the embedding level. By context-subtractive query steering, GRAIL excels at compositional cross-modal reasoning, while additive embedding updates show strength on localized information aggregation. By dynamically routing queries based on task type, our Hybrid Framework achieves a 40.3\% macro-averaged performance gain on MultimodalQA. Extensive evaluations demonstrate that sequential GRAIL retrieves in a superior, noise-resilient manner, significantly expanding the search horizon through iterative gap-aware optimization. |
| title | Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2605.28641 |