Metraipolis › archive › dc_d10116d04741

Breaking the Tie: A Cluster-Aware Routing Framework for Large Language Models — arXiv:2610.05982

Source
arXiv
Original
https://arxiv.org/abs/2610.05982
License (as archived)
CC0-1.0
Archived
2026-10-06T05:28:28.024Z
SHA-256 of the archived text
13967e09b086d589f2d7b25d8f9919d5e3fde56e22c06e40a4b35988c048f2b3
Stable URL of this page
https://metraipolis.com/documento/dc_d10116d04741

How to cite

Breaking the Tie: A Cluster-Aware Routing Framework for Large Language Models — arXiv:2610.05982 — arXiv, https://arxiv.org/abs/2610.05982 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d10116d04741 (Metraipolis archive dc_d10116d04741, sha256 13967e09b086d589f2d7b25d8f9919d5e3fde56e22c06e40a4b35988c048f2b3)
BibTeX
@misc{dcd10116d04741,
  title = {Breaking the Tie: A Cluster-Aware Routing Framework for Large Language Models — arXiv:2610.05982},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.05982},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d10116d04741; sha256 13967e09b086d589f2d7b25d8f9919d5e3fde56e22c06e40a4b35988c048f2b3},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.05982. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

With the rapid development of artificial intelligence, the emergence of various Large Language Models (LLMs) has created a rich model ecosystem. However, this also brings a key challenge: how to select the optimal model for a specific user query. LLM routing addresses this need by dynamically assigning queries to the most suitable expert in the pool of candidate models. However, existing routing frameworks often simplify this process to a standard classification task; thus, a critical vulnerability is exposed when multiple candidate models correctly answer the same query. We formalize this capability overlap as routing noise, which misleads the router with arbitrarily correct candidate models, ultimately leading to routing collapse (a severe decline in generalization ability on unseen tasks). To address this problem, we propose a novel Cluster-Aware Soft-Labeling Routing (CASLR) framework. CASLR shifts the evaluation paradigm from the success of a single query to macro-domain consensus by replacing traditional one-hot vectors with a masked softmax mechanism. Specifically, for experts who answer incorrectly, we penalize their target probability to zero; for the remaining candidates, we directly compute continuous fine-grained soft labels based on their global clustering utility scores. We then use these refined soft labels to supervise a lightweight router. Specifically, the framework not only demonstrates superior accuracy on multiple benchmarks, but also outperforms Llama-3.3-70B-Instruct by 7.80% in overall average performance. Furthermore, the extremely low routing inference latency of only 1.13s further confirms that CASLR can achieve efficient system scheduling with almost zero additional overhead, while ensuring high response quality.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.05982 — 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 2026-10-06T05:24:20.354Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).