$T^2$ of Thoughts：温度树激发大型语言模型的推理能力
$T^2$ of Thoughts: Temperature Tree Elicits Reasoning in Large Language Models
AI总结：
本文提出 $T^2oT$ 启发式提示方法，通过温度树动态调整温度等搜索参数，在不增加计算开销的前提下提升 LLM 的推理、决策与文本生成表现。
摘要：
大型语言模型（LLM）已成为人工智能领域的强大工具，尤其是在复杂决策场景中，但其静态的问题解决策略往往限制了它们对动态环境的适应能力。我们通过一种称为 $T^2$ of Thoughts（$T^2oT$）的启发式算法，探索利用温度树（Temperature Tree，$T^2$）提示来增强 LLM 的推理能力。主要重点是通过动态调整搜索参数（尤其是温度）来增强决策过程，从而在不增加计算需求的情况下提高准确性。我们通过实验验证，混合式 $T^2oT$ 方法在单解准确性、多解生成和文本生成质量方面均带来提升。研究结果表明，尽管基于温度的动态搜索深度调整可能产生好坏参半的结果，但固定搜索深度结合 $T^2oT$ 的自适应能力，可提供更可靠、更通用的问题解决策略。本工作凸显了未来优化算法与基础语言模型交互的潜力，尤其通过我们为 Game of 24 和 Creative Writing 任务所做的开发得到说明。
英文摘要：
Large Language Models (LLMs) have emerged as powerful tools in artificial intelligence, especially in complex decision-making scenarios, but their static problem-solving strategies often limit their adaptability to dynamic environments. We explore the enhancement of reasoning capabilities in LLMs through Temperature Tree ($T^2$) prompting via a heuristic algorithm, termed as $T^2$ of Thoughts ($T^2oT$). The primary focus is on enhancing decision-making processes by dynamically adjusting search parameters, especially temperature, to improve accuracy without increasing computational demands. We empirically validate that our hybrid $T^2oT$ approach yields enhancements in, single-solution accuracy, multi-solution generation and text generation quality. Our findings suggest that while dynamic search depth adjustments based on temperature can yield mixed results, a fixed search depth, when coupled with adaptive capabilities of $T^2oT$, provides a more reliable and versatile problem-solving strategy. This work highlights the potential for future explorations in optimizing algorithmic interactions with foundational language models, particularly illustrated by our development for the Game of 24 and Creative Writing tasks.
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。
