Metraipolis › archive › dc_c3cb0e2bb934

G2MAF: Test-Time Gradient Guidance for Multi-Agent Flow Policies — arXiv:2609.31286

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.31286
License (as archived)
CC0-1.0
Archived
2026-10-08T01:38:40.562Z
SHA-256 of the archived text
9b3e81258cb8cb8fedc3aaf4456123576428fd9232caa58ce7f098cbfa83b223
Stable URL of this page
https://metraipolis.com/documento/dc_c3cb0e2bb934

How to cite

G2MAF: Test-Time Gradient Guidance for Multi-Agent Flow Policies — arXiv:2609.31286 — arXiv (en), https://arxiv.org/abs/2609.31286 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c3cb0e2bb934 (Metraipolis archive dc_c3cb0e2bb934, sha256 9b3e81258cb8cb8fedc3aaf4456123576428fd9232caa58ce7f098cbfa83b223)
BibTeX
@misc{dcc3cb0e2bb934,
  title = {G2MAF: Test-Time Gradient Guidance for Multi-Agent Flow Policies — arXiv:2609.31286},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.31286},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c3cb0e2bb934; sha256 9b3e81258cb8cb8fedc3aaf4456123576428fd9232caa58ce7f098cbfa83b223},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31286. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Offline multi-agent reinforcement learning (MARL) learns cooperative policies from fixed datasets without further environment interaction and a learned policy is frozen at deployment. Such a frozen policy typically proposes a single joint action and executes it directly at deployment time. However, this one-shot deployment often commits to a suboptimal proposal, even when better nearby alternatives remain consistent with the behavior data. To address this issue, we propose Gradient Guided Multi Agent Flow (G2MAF), a refinement framework for optimizing joint policies at test-time. G2MAF applies one globally normalized, projected critic gradient to guide and coordinate all agents' corrections while keeping the action both feasible and close to the frozen policy proposal. Across 24 MPE and SMAC settings, its canonical variant improves 20 frozen settings, with mean relative gains of 9.2% on MPE and 8.9% on SMAC, with model inference latency increased by about 6% only.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-28 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.31286 — 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 2026-10-08T01:12:14.368Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).