Metraipolis › archive › dc_d30dae236d21
Agent-Conditioned Routing: Diverse Reasoning Paths Within One Frozen Model — arXiv:2610.37375
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.37375
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-28T03:02:05.255Z
- SHA-256 of the archived text
9b057b313e1eb0a78b38a5f5cb8543be33448f6aacd71d7070aef19653d34478- Stable URL of this page
https://metraipolis.com/documento/dc_d30dae236d21
How to cite
Agent-Conditioned Routing: Diverse Reasoning Paths Within One Frozen Model — arXiv:2610.37375 — arXiv (en), https://arxiv.org/abs/2610.37375 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d30dae236d21 (Metraipolis archive dc_d30dae236d21, sha256 9b057b313e1eb0a78b38a5f5cb8543be33448f6aacd71d7070aef19653d34478)
BibTeX
@misc{dcd30dae236d21,
title = {Agent-Conditioned Routing: Diverse Reasoning Paths Within One Frozen Model — arXiv:2610.37375},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.37375},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d30dae236d21; sha256 9b057b313e1eb0a78b38a5f5cb8543be33448f6aacd71d7070aef19653d34478},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.37375. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Generating several solutions can expose alternative methods, but additional trajectories often paraphrase one dominant approach. We investigate whether the conditional capacity of a frozen sparse mixture-of-experts (MoE) model can provide such diversity from one problem prompt, without user-assigned roles, prescribed methods, fine-tuning, or model copies. Agent-Conditioned Routing (ACR) assigns $m$ parallel trajectories persistent, untrained shared/private expert-eligibility masks while preserving the learned, input-dependent token router. On a screened 126-question multi-mode mathematics and physics benchmark, with model judges reading the reported derivations in full, we count the number of problems yielding at least two distinct primary methods among soft-correct solutions in four trajectories. ACR reaches approximately twice as many such problems as the strongest temperature setting under GLM 5.3 (37.3 versus 18.7), and 1.67 times as many under GPT-5.6 Sol (45.00 versus 27.00), at matched output budgets. A blinded four-expert audit on a stratified sample of 30 problems also finds more multi-method solutions with ACR on both backbones, with lower correctness. Matched-eligibility ablations show modest multi-method count advantages for ACR, while token-resampled masks attain higher correctness. An informed sequential prompt elicits more diversity still, using previous solutions and explicit requests for alternative methods. ACR adds about 160 KiB of mask state. In a three-problem microbenchmark capped at 1,024 tokens, its latency and peak VRAM usage remain within 3% of batched temperature sampling. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG); Multiagent Systems (cs.MA). Pagina del lavoro: https://arxiv.org/abs/2610.37375 — 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 2026-10-28T03:01:03.538Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).