@inproceedings{liu-etal-2026-metascale,
    title = "{M}eta{S}cale: Test-Time Scaling with Evolving Meta-Thoughts",
    author = "Liu, Qin  and
      Zhou, Wenxuan  and
      Xu, Nan  and
      Huang, James Y.  and
      Wang, Fei  and
      Zhang, Sheng  and
      Poon, Hoifung  and
      Chen, Muhao",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.574/",
    pages = "11828--11842",
    ISBN = "979-8-89176-395-1",
    abstract = "One critical challenge for large language models (LLMs) in making complex reasoning is their reliance on matching reasoning patterns from training data, instead of proactively selecting the most appropriate cognitive strategy to solve a given task. Existing approaches impose fixed cognitive structures that enhance performance in specific tasks but lack adaptability across diverse scenarios. To address this limitation, we introduce MetaScale, a test-time scaling framework based on meta-thoughts, i.e., adaptive thinking strategies tailored to each task. MetaScale initializes a pool of candidate meta-thoughts, then iteratively selects and evaluates them using a multi-armed bandit algorithm with upper confidence bound selection, guided by a reward model. To further enhance adaptability, a genetic algorithm evolves high-reward meta-thoughts, refining and extending the strategy pool over time. By dynamically proposing and optimizing meta-thoughts at inference time, MetaScale improves both accuracy and generalization across a wide range of tasks. Experimental results demonstrate that MetaScale consistently outperforms standard inference approaches, achieving an 11{\%} performance gain in win rate on Arena-Hard with GPT-4o, improving from 82.14{\%} to 93.14{\%} against GPT-4. Notably, MetaScale scales more effectively with increasing sampling budgets and produces more structured, expert-level responses."
}