Metraipolis › archive › dc_52223ade9ee7
ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models — arXiv:2610.07803
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.07803
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:19:10.506Z
- SHA-256 of the archived text
91961241e2174cea20c05801cea87e67efc8ef6e7c7f3fae2041b7d69a37c957- Stable URL of this page
https://metraipolis.com/documento/dc_52223ade9ee7
How to cite
ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models — arXiv:2610.07803 — arXiv (en), https://arxiv.org/abs/2610.07803 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_52223ade9ee7 (Metraipolis archive dc_52223ade9ee7, sha256 91961241e2174cea20c05801cea87e67efc8ef6e7c7f3fae2041b7d69a37c957)
BibTeX
@misc{dc52223ade9ee7,
title = {ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models — arXiv:2610.07803},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.07803},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_52223ade9ee7; sha256 91961241e2174cea20c05801cea87e67efc8ef6e7c7f3fae2041b7d69a37c957},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07803. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Small reasoning models (SRMs) have shown strong performance on complex reasoning tasks by generating extended chain-of-thought trajectories, but they often fail to recover once their reasoning enters an erroneous path. Existing test-time fusion methods rely on local fusion signals to determine when to trigger fusion, which can be misled by transient uncertainty fluctuations and may reinforce unstable reasoning trajectories. We propose ThinkFuse, a training-free test-time fusion framework that selectively intervenes in unreliable reasoning segments. ThinkFuse compares segment-level uncertainty shifts with trajectory-level uncertainty trends to identify unstable reasoning points and fuse auxiliary reasoning paths into the primary model's trajectory. Extensive experiments demonstrate that ThinkFuse outperforms baselines on mathematical and knowledge-intensive reasoning benchmarks, with consistent gains across model-family combinations, and remains robust with a smaller primary model. Our analysis shows that ThinkFuse requires fewer fusion triggers and generates fewer tokens, highlighting the efficiency of selective triggering. Our code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.07803 — 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 2026-10-07T11:15:24.558Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).