Question's Gambit is introduced, a first-move retrieval module that decomposes the question into a set of clues, reformulates them into complementary searches, consolidates the retrieved results, and reranks the candidate pool before the agent begins its iterative search-and-reasoning process.
Abstract
Deep research agents answer complex questions through iterative loops of searching, reading, and reasoning. Recent work on reasoning-intensive benchmarks such as BrowseComp-Plus shows that well-configured lexical retrieval can surface high-quality evidence, yet agents may still fail to connect documents carrying evidence to the gold documents. We identify a deep research agent's first retrieval move as an important design decision for this setting. We introduce Question's Gambit, a first-move retrieval module that decomposes the question into a set of clues, reformulates them into complementary searches, consolidates the retrieved results, and reranks the candidate pool before the agent begins its iterative search-and-reasoning process. This produces an opening context designed to support both clue aggregation and final-answer verification. We further evaluate on MultiHop-RAG to test whether these benefits transfer beyond BrowseComp-Plus to a more conventional multi-hop question structure. Experiments on BrowseComp-Plus show that Question's Gambit improves retrieval recall and downstream agent accuracy over strong baselines, improving answer accuracy from 83.1% to 90.5% with gpt-5.5 over Pi-Serini, the strongest reported agentic baseline. Our results confirm that effective agentic deep research depends not only on the tools available inside the loop, but also on the quality of the first move. We published our implementation publicly at https://github.com/radinhamidi/Question-s-Gambit.
IndexAct is proposed, an interface for Index-Native Corpus Interaction that separates candidate-set refinement from text inspection, and achieves higher evidence coverage with a smaller average live context than terminal-based corpus interfaces, and maintains answer accuracy as the corpus expands.
Deogyong Kim, Sunghwan Kim, Sangam Lee et al.· 0 citations
Retrieval-Grounded Voting (RGV), which scores each rollout by the lexical overlap between its final answer and the documents it retrieved, consistently outperforms confidence-based voting and identifies the underlying failure reason as copy inflation.
Hyunho Kook, Junhyuk So, Tianyu Fu et al.· 0 citations
HypoSearch is proposed, which generates lightweight hypotheses as soft search hints, explores them through bounded independent branches, and compares branch-level evidence before commitment, and consistently outperforms single-trajectory search and standard parallel baselines.
Ruo-Chen Zhou, Zheng-Zong Chen, Luan Zhang et al.· 1 citation
ITER, an agent interaction-aware dense retriever trained using agent trajectory learning signals, is introduced andlations show that structured interaction history and pre-search reasoning provide complementary retrieval context, while previously visited and useful documents provide the strongest trajectory-relative su...
Hao-Dong Chen, Shuai Wang, Yu Yin et al.· 0 citations
Search agents extend large reasoning models with multi-round access to external knowledge, yet most existing systems largely conflate retrieving evidence with consuming it. Retrieved documents are commonly incorporated according to fixed or rank-based policies, despite the fact that evidence utility is conditioned on t...
Jia-Jin Liu, Chen-Xi Xu, Dong-Zhe Fan et al.· 0 citations
A task-adaptive two-layer methodology combining a compact, reviewed knowledge layer for orientation and cross-source synthesis with a versioned evidence layer for exact lookup and claim support is presented.
A. Volkov, Taras Pustovoy, Dmitriy Istomin et al.· Machine Learning and Knowled...· 0 citations