Test-Time Scaling (TTS) is a promising approach to progressively elicit the model's intelligence during inference. Recently, training-based TTS methods, such as continued reinforcement learning (RL), have further surged in popularity, while training-free TTS methods are gradually fading from prominence. However, the additional computation overhead of training amplifies the burden on test-time scaling. In this paper, we focus on training-free TTS methods for reasoning. We first design Conditional Step-level Self-refinement, a fine-grained sequential scaling method guided by process verification. On top of its effectiveness, we further combine it with other classical parallel scaling methods at the step level, to introduce a novel inference paradigm called Hybrid Test-Time Scaling. Extensive experiments on five instruction-tuned LLMs across different scales (3B-14B) and families demonstrate that hybrid strategy incorporating various training-free TTS methods at a fine granularity has considerable potential for expanding the reasoning performance boundaries of LLMs.
@article{arxiv.2507.15512,
title = {Step-level Verifier-guided Hybrid Test-Time Scaling for Large Language Models},
author = {Kaiyan Chang and Yonghao Shi and Chenglong Wang and Hang Zhou and Chi Hu and Xiaoqian Liu and Yingfeng Luo and Yuan Ge and Tong Xiao and Jingbo Zhu},
journal= {arXiv preprint arXiv:2507.15512},
year = {2025}
}
Comments
Accepted by EMNLP 2025. Code: https://github.com/Lucky-259/Hybrid_TTS