Metraipolis › archive › dc_ff442a06ce48

RSI-Router: Evolving Subtask-Level LLM Routing and Skills for Cost-Efficient Agents — arXiv:2609.34712

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34712
License (as archived)
CC0-1.0
Archived
2026-10-07T20:58:05.702Z
SHA-256 of the archived text
45eb106574847caa5c5fab267b5469ae9f4877ca9938068d6e140800fe1a883b
Stable URL of this page
https://metraipolis.com/documento/dc_ff442a06ce48

How to cite

RSI-Router: Evolving Subtask-Level LLM Routing and Skills for Cost-Efficient Agents — arXiv:2609.34712 — arXiv (en), https://arxiv.org/abs/2609.34712 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ff442a06ce48 (Metraipolis archive dc_ff442a06ce48, sha256 45eb106574847caa5c5fab267b5469ae9f4877ca9938068d6e140800fe1a883b)
BibTeX
@misc{dcff442a06ce48,
  title = {RSI-Router: Evolving Subtask-Level LLM Routing and Skills for Cost-Efficient Agents — arXiv:2609.34712},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34712},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ff442a06ce48; sha256 45eb106574847caa5c5fab267b5469ae9f4877ca9938068d6e140800fe1a883b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34712. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Practical deployment of large language model (LLM) agents requires strong task performance at affordable inference cost. For long-horizon agentic tasks, this performance-cost trade-off can be improved through within-task large-small model collaboration, as smaller models can handle some stages even when they cannot solve the full task. In this paper, we introduce RSI-router, a routing framework that constructs subtask-level model assignments and model-specific skills through recursive self-improvement over accumulated experience. Each iteration consists of four stages: Subtask Mining derives subtask definitions and identification rules from training trajectories; Routing Strategy Evolution proposes and evaluates diverse model assignments; Model-Specific Skill Evolution compares routed and large-model-only trajectories to diagnose failures and develop reusable execution skills; and Pareto-Optimal Router Selection updates the Pareto population using historical and newly generated routers while retaining dominated routers as experience for subsequent evolution. Routing between DeepSeek-V4.1-Flash and Qwen3.5-9B, RSI-router consistently surpasses the DeepSeek-only baseline at roughly half the inference cost (48.3%) across five agentic benchmarks. In particular, on ALFWorld, ScienceWorld, and WebShop, it cuts inference cost by 74.7-82.2% while simultaneously improving performance; on Terminal-Bench 2.0, it achieves a 16.7% relative performance gain at 18.0% lower cost. Moreover, RSI-router establishes a stronger performance--cost Pareto frontier than 9 routing methods.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.34712 — 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 2026-10-07T20:29:01.638Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).