ML Anthology

What Are Step-Level Reward Models Rewarding? Counterintuitive Findings from MCTS-Boosted Mathematical Reasoning

Yiran Ma, Zui Chen, Tianqiao Liu, Mi Tian, Zhuo Liu, Zitao Liu, Weiqi Luo
AAAI 2025 pp. 24812-24820
doi:10.1609/AAAI.V39I23.34663 /aaai/2025/ma2025aaai-step/
AAAI Semantic Scholar

Cite

Text

Ma et al. "What Are Step-Level Reward Models Rewarding? Counterintuitive Findings from MCTS-Boosted Mathematical Reasoning." AAAI Conference on Artificial Intelligence, 2025. doi:10.1609/AAAI.V39I23.34663

Markdown

[Ma et al. "What Are Step-Level Reward Models Rewarding? Counterintuitive Findings from MCTS-Boosted Mathematical Reasoning." AAAI Conference on Artificial Intelligence, 2025.](https://mlanthology.org/aaai/2025/ma2025aaai-step/) doi:10.1609/AAAI.V39I23.34663

BibTeX

@inproceedings{ma2025aaai-step,
  title     = {{What Are Step-Level Reward Models Rewarding? Counterintuitive Findings from MCTS-Boosted Mathematical Reasoning}},
  author    = {Ma, Yiran and Chen, Zui and Liu, Tianqiao and Tian, Mi and Liu, Zhuo and Liu, Zitao and Luo, Weiqi},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2025},
  pages     = {24812-24820},
  doi       = {10.1609/AAAI.V39I23.34663},
  url       = {https://mlanthology.org/aaai/2025/ma2025aaai-step/}
}