更宽还是更深？利用自适应分支树搜索扩展大语言模型推理时计算
Wider or Deeper? Scaling LLM Inference-Time Compute with Adaptive Branching Tree Search
AI总结：
本文提出自适应分支蒙特卡洛树搜索（AB-MCTS），通过外部反馈动态决定扩展新候选响应或细化已有响应，在复杂编程任务上优于重复采样和标准MCTS，有效提升推理时扩展性能。
摘要：
近期进展表明，增加推理时计算可以显著提升大语言模型（LLMs）的推理能力。尽管重复采样（即生成多个候选输出）是一种非常有效的策略，但它并未利用外部反馈信号进行细化，而这些信号在编程等任务中通常是可获得的。在本工作中，我们提出了自适应分支蒙特卡洛树搜索（AB-MCTS），这是一种新颖的推理时框架，通过有原则的多轮探索与利用来推广重复采样。在搜索树的每个节点上，AB-MCTS 根据外部反馈信号动态决定是“走得更宽”，即扩展新的候选响应，还是“走得更深”，即重新审视已有响应。我们使用前沿模型在复杂的编程和工程任务上评估了该方法。实验结果表明，AB-MCTS 一致地优于重复采样和标准 MCTS，凸显了将大语言模型的响应多样性与多轮解细化相结合对于有效推理时扩展的重要性。代码可在 https://github.com/SakanaAI/treequest 获取。
英文摘要：
Recent advances demonstrate that increasing inference-time computation can significantly boost the reasoning capabilities of large language models (LLMs). Although repeated sampling (i.e., generating multiple candidate outputs) is a highly effective strategy, it does not leverage external feedback signals for refinement, which are often available in tasks like coding. In this work, we propose Adaptive Branching Monte Carlo Tree Search (AB-MCTS), a novel inference-time framework that generalizes repeated sampling with principled multi-turn exploration and exploitation. At each node in the search tree, AB-MCTS dynamically decides whether to "go wider" by expanding new candidate responses or "go deeper" by revisiting existing ones based on external feedback signals. We evaluate our method on complex coding and engineering tasks using frontier models. Empirical results show that AB-MCTS consistently outperforms both repeated sampling and standard MCTS, underscoring the importance of combining the response diversity of LLMs with multi-turn solution refinement for effective inference-time scaling. Code is available at https://github.com/SakanaAI/treequest .
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
