Metraipolis › archive › dc_05ddefa03b82

R$^2$Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization Loop — arXiv:2610.22740

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.22740
License (as archived)
CC0-1.0
Archived
2026-10-19T08:42:13.920Z
SHA-256 of the archived text
7fe3c13b3d87d3827d1aab177b06f32aa5573f8a0ba60656cec59043a22b4475
Stable URL of this page
https://metraipolis.com/documento/dc_05ddefa03b82

How to cite

R$^2$Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization Loop — arXiv:2610.22740 — arXiv (en), https://arxiv.org/abs/2610.22740 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_05ddefa03b82 (Metraipolis archive dc_05ddefa03b82, sha256 7fe3c13b3d87d3827d1aab177b06f32aa5573f8a0ba60656cec59043a22b4475)
BibTeX
@misc{dc05ddefa03b82,
  title = {R$^2$Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization Loop — arXiv:2610.22740},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.22740},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_05ddefa03b82; sha256 7fe3c13b3d87d3827d1aab177b06f32aa5573f8a0ba60656cec59043a22b4475},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22740. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In recent years, agentic orchestration has been widely applied to solve complex human tasks. However, existing orchestration learning still faces key challenges, including hand-designed curricula, advantage collapse when a task's rollouts all reach the same outcome, and unreliable evaluation signals that judge each skill by one noisy score. To address these challenges, we propose R$^2$Orch, a new paradigm of Recursive Self-Play Orchestration: an inner loop of self-play that improves the orchestrator on self-chosen tasks, and an outer loop that improves how the inner loop improves. To support this paradigm, we introduce Recursive Learnability, a training objective that rewards a Proposer for choosing tasks whose rollouts keep reaching different outcomes as skills are withdrawn. Built on this objective, we further propose Skill Withdrawal, in which the outer loop withdraws each skill once a controlled evaluation within self-play shows that the orchestrator performs as well without it. Moreover, the two loops form one cycle: failures are distilled into skills, skills into parameters, and withdrawals lead the Proposer to new failures, so the orchestrator moves toward zero-skill deployment. Experimental results on twelve datasets show that R$^2$Orch consistently outperforms baselines across question answering, math, code, and interactive tasks.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.22740 — 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 2026-10-19T08:39:10.031Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).