Metraipolis › archive › dc_fa59ed904c95
From Delegation to Coordination: Learning Efficient Multi-Agent Deep Research — arXiv:2611.17041
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.17041
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-14T02:39:05.573Z
- SHA-256 of the archived text
31caa182b57f0861c815a184f928cfb4ae2645eb6438b92024ab4ebef31a08bf- Stable URL of this page
https://metraipolis.com/documento/dc_fa59ed904c95
How to cite
From Delegation to Coordination: Learning Efficient Multi-Agent Deep Research — arXiv:2611.17041 — arXiv (en), https://arxiv.org/abs/2611.17041 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fa59ed904c95 (Metraipolis archive dc_fa59ed904c95, sha256 31caa182b57f0861c815a184f928cfb4ae2645eb6438b92024ab4ebef31a08bf)
BibTeX
@misc{dcfa59ed904c95,
title = {From Delegation to Coordination: Learning Efficient Multi-Agent Deep Research — arXiv:2611.17041},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.17041},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fa59ed904c95; sha256 31caa182b57f0861c815a184f928cfb4ae2645eb6438b92024ab4ebef31a08bf},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.17041. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-agent deep-research systems use delegation to parallelize search and keep long tool traces out of the main agent's context. Yet delegation helps only when task allocation and subtask execution are coordinated: the main agent must issue useful subtasks, while subagents must return reliable evidence. Existing training either optimizes only the main agent or broadcasts the same team advantage to all agents, which can reward misleading subagent reports and penalize useful ones. We introduce LEAD-RL (Learning Effective Agent Delegation with Reinforcement Learning), a joint main-agent-subagent training framework with contribution-aware credit assignment. A training-time rubric evaluates the research utility of each subagent report and gates its inherited team advantage, while final-answer correctness remains the only team reward; no token, step, or speedup reward is used. Across five benchmarks and two model families, LEAD-RL improves answer quality. On BrowseComp, it improves accuracy by 2.49 percentage points and reduces critical-path cost by 24.5% over joint training with shared team credit. During validation, accuracy increases by 6.78 points at approximately constant critical-path cost. These results show that contribution-aware credit assignment improves the quality-efficiency trade-off of multi-agent deep research. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno). Categorie arXiv: Multiagent Systems (cs.MA) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.17041 — 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 2026-11-14T02:39:04.727Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).