Metraipolis › archive › dc_502bd3e90771
SEPAL: Separated Expert Pairs with Answer-Level Fusion for Reliable LLM Collaboration — arXiv:2609.39645
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.39645
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T12:45:40.072Z
- SHA-256 of the archived text
9ea85e2154e718b6a853f50be4e18f1f52796056bfc53125361523eb75b98e63- Stable URL of this page
https://metraipolis.com/documento/dc_502bd3e90771
How to cite
SEPAL: Separated Expert Pairs with Answer-Level Fusion for Reliable LLM Collaboration — arXiv:2609.39645 — arXiv (en), https://arxiv.org/abs/2609.39645 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_502bd3e90771 (Metraipolis archive dc_502bd3e90771, sha256 9ea85e2154e718b6a853f50be4e18f1f52796056bfc53125361523eb75b98e63)
BibTeX
@misc{dc502bd3e90771,
title = {SEPAL: Separated Expert Pairs with Answer-Level Fusion for Reliable LLM Collaboration — arXiv:2609.39645},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.39645},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_502bd3e90771; sha256 9ea85e2154e718b6a853f50be4e18f1f52796056bfc53125361523eb75b98e63},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39645. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-agent collaboration lets large language models (LLMs) improve question answering through deliberation and feedback. Yet shared discussion couples correction with exposure to the same mistakes, which can erode the diversity needed for voting. Self-consistency offers sampling diversity without feedback, while single-pair Actor-Critic collaboration refines only one candidate. We introduce SEPAL, which assigns three private Actor-Critic teams to direct reasoning, evidence grounding, and verification. Role-specific training gives the teams different reasoning objectives beyond sampling variation. Each Critic guides revisions within its own team, preventing feedback from carrying errors across candidates. Once revision ends, majority voting combines only the final answers, keeping the reasoning histories separate until the decision. Across five open-weight backbones and five question-answering benchmarks, SEPAL improves mean accuracy by 1.81 percentage points over a matched single Actor-Critic pair, with improvements across all five backbones. Code is available at 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); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.39645 — 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 2026-10-07T12:14:14.131Z 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).