Abstract:Large reasoning models (LRMs) combined with retrieval-augmented generation (RAG) have enabled deep research agents capable of multi-step reasoning with external knowledge retrieval. However, we find that existing approaches rarely demonstrate test-time search scaling. Methods that extend reasoning through single-query sequential search suffer from limited evidence coverage, while approaches that generate multiple independent queries per step often lack structured aggregation, hindering deeper sequential reasoning. We propose a hybrid search strategy to address these limitations. We introduce HybridDeepSearcher, a structured search agent that integrates parallel query expansion with explicit evidence aggregation before advancing to deeper sequential reasoning. To supervise this behavior, we introduce HDS-QA, a novel dataset that guides models to combine broad parallel search with structured aggregation through supervised reasoning-query0retrieval trajectories containing parallel sub-queries. Across five benchmarks, HybridDeepSearcher significantly outperforms the state-of-the-art, improving F1 scores by +15.9 on FanOutQA and +9.2 on a subset of BrowseComp. Further analysis shows its consistent test-time search scaling: performance improves as additional search turns or calls are allowed, while competing methods plateau.
| Comments: | Accepted to ICLR 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2508.19113 [cs.AI] |
| (or arXiv:2508.19113v3 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2508.19113 arXiv-issued DOI via DataCite |
From: Dayoon Ko [view email]
[v1]
Tue, 26 Aug 2025 15:15:17 UTC (3,565 KB)
[v2]
Tue, 24 Feb 2026 09:32:16 UTC (2,342 KB)
[v3]
Mon, 25 May 2026 15:33:03 UTC (4,091 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。