Abstract
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 https://github.com/js-lee-AI/ThinkFuse.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Kang, M., Lee, J., Seo, J., & Lim, H. (2026). ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models. https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models
MLA 9
Kang, Myunghoon, et al. "ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models." https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models.
Chicago (author–date)
Kang, Myunghoon, Jungseob Lee, Jaehyung Seo, and Heuiseok Lim. 2026. "ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models." https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models.
Harvard
Kang, M., Lee, J., Seo, J. and Lim, H. (2026) 'ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models', Available at: https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models.
Vancouver
Kang M, Lee J, Seo J, Lim H. ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models. https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models
IEEE
M. Kang, J. Lee, J. Seo, and H. Lim, "ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models," https://omanscience.com/en/articles/thinkfuse-trajectory-aware-test-time-fusion-for-small-reasoning-models.