Metraipolis › archive › dc_0f375bbaf260

DAGent: Evaluate-then-Grow Planning for Deep Research Agents — arXiv:2609.39154

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39154
License (as archived)
CC0-1.0
Archived
2026-10-07T12:45:00.624Z
SHA-256 of the archived text
71d0fa13d36eac243bbf54a11d382343e7a4e384214c6de69295cd3e2a49680d
Stable URL of this page
https://metraipolis.com/documento/dc_0f375bbaf260

How to cite

DAGent: Evaluate-then-Grow Planning for Deep Research Agents — arXiv:2609.39154 — arXiv (en), https://arxiv.org/abs/2609.39154 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0f375bbaf260 (Metraipolis archive dc_0f375bbaf260, sha256 71d0fa13d36eac243bbf54a11d382343e7a4e384214c6de69295cd3e2a49680d)
BibTeX
@misc{dc0f375bbaf260,
  title = {DAGent: Evaluate-then-Grow Planning for Deep Research Agents — arXiv:2609.39154},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39154},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0f375bbaf260; sha256 71d0fa13d36eac243bbf54a11d382343e7a4e384214c6de69295cd3e2a49680d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39154. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Deep research tasks require agents to navigate large knowledge spaces, synthesize evidence across many sources, and adapt their plans as findings emerge. Directed acyclic graph (DAG)-based multi-agent systems suit this setting because they support parallel execution and isolate each sub-task within a focused dependency context. Yet existing DAG-based agents instantiate a task-level plan before execution and repair the graph only after failures or missing evidence are observed. This Plan-then-Patch strategy is brittle for deep research: the system commits most strongly when its evidence is weakest, and later revisions waste computation on branches that should not have been planned. We propose DAGent, a DAG-based multi-agent framework with Evaluate-then-Grow incremental planning: an Orchestrator grows the task graph one batch at a time, conditioning each expansion on confidence and uncertainty signals from completed nodes. A hierarchical context layer propagates compact QueryDocs by default while preserving full execution traces for on-demand recall. The recorded DAG topology admits structural RL signals that outcome-only recipes cannot define; DAGRPO, a GRPO adaptation, injects topology-conditioned credit on Executor rollouts and a structural compliance regularization on Orchestrator plans. Across BrowseComp-Plus, GAIA, and xbench-DeepSearch, DAGent surpasses the strongest open-source baseline by 5.3 / 5.8 / 2.0 points at the Qwen3-235B-A22B scale, and the lead replicates across four open-source backbones and extends to GPT-5 at 327K context. At the Qwen3-8B scale, DAGRPO improves over a same-budget outcome-only GRPO baseline by 3.0 average Pass@1 points. A same-architecture comparison shows that evidence-conditioned planning reaches higher accuracy at lower per-task token, tool-call, and step footprints than its Plan-then-Patch counterpart. Code: this https URL

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.39154 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T12:14:14.030Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).