Metraipolis › archive › dc_17d5bbd86fc9

SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology — arXiv:2609.02292

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.02292
License (as archived)
CC0-1.0
Archived
2026-10-17T03:26:23.220Z
SHA-256 of the archived text
4a6487ee3c097aa50f4ffa81e0f96603dace5749e6e7e459539bf4670a7fa3d8
Stable URL of this page
https://metraipolis.com/documento/dc_17d5bbd86fc9

How to cite

SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology — arXiv:2609.02292 — arXiv (en), https://arxiv.org/abs/2609.02292 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_17d5bbd86fc9 (Metraipolis archive dc_17d5bbd86fc9, sha256 4a6487ee3c097aa50f4ffa81e0f96603dace5749e6e7e459539bf4670a7fa3d8)
BibTeX
@misc{dc17d5bbd86fc9,
  title = {SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology — arXiv:2609.02292},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.02292},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_17d5bbd86fc9; sha256 4a6487ee3c097aa50f4ffa81e0f96603dace5749e6e7e459539bf4670a7fa3d8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02292. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

The rapid proliferation of large language models (LLMs) and the growing diversity of their applications presents a unique optimization opportunity: selecting the right model for the task, while optimizing for speed, cost, and quality at a per-task level. However, inference endpoints can vary widely in quality, price, latency, context support, tool use, domain expertise, and reasoning behavior. This heterogeneity makes manual heuristics difficult to maintain and unlikely to achieve consistently favorable speed--cost--quality trade-offs on their own. We introduce \router{}, a lightweight GLiClass-based router that assigns a suitability score to each inference-time model label without autoregressive generation. The released 0.6B-parameter checkpoint combines a Qwen3 decoder with a shallow bidirectional scorer. Its decoder-KV execution path preserves a text-only key--value cache across a session, encodes only new dialogue turns, and evaluates transient candidate-label tokens without adding them to the persistent cache. The same checkpoint also predicts task type, difficulty, reasoning mode, and expected output length, and supports custom zero-shot labels. For task generation, we construct a task ontology with 23 families, 115 task types, 345 routable subtypes, 1,173 synthetic examples, and an orthogonal axis of 30 domains. Using this structure, we generate 150,000 verifier-scored tasks and 15,000 open-ended tasks. We then train the Qwen3 decoder on these tasks, while explicitly separating learned request prediction from per-task policies for attributes such as eligibility, cost, cache reuse, safety, and sovereignty. Across six LiveBench subsets, the router outperforms the mean candidate; on the selected 1,000-task subset, it achieves an aggregate top-1 score of 0.707 versus 0.696 for the strongest fixed model, with benchmark-dependent gains.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.02292 — 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 2026-10-17T03:13:12.438Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).