思維狀態模板：樹狀思維結構化行動模板

STATe-of-Thoughts: Structured Action Templates for Tree-of-Thoughts

February 15, 2026

作者: Zachary Bamberger, Till R. Saenger, Gilad Morad, Ofra Amir, Brandon M. Stewart, Amir Feder

cs.AI

摘要

諸如最佳N取樣（Best-of-N）與思維樹（Tree-of-Thoughts）等推論時計算（ITC）方法，旨在生成兼具高質量與多樣性的輸出候選方案，但其採用高溫取樣策略往往難以實現實質性的輸出多樣性。此外，現有ITC方法對推理過程的控制能力有限，這也制約了其可解釋性。我們提出「結構化思維」（STATe）這一可解釋的ITC方法，透過對高層次推理模式進行搜索來解決上述問題。STATe以離散且可解釋的文本干預取代隨機取樣：控制器選擇編碼高層次推理決策的動作，生成器根據這些選擇產生推理步驟，評估器則對候選方案評分以引導搜索。此結構化方法具備三大優勢：首先，動作引導的文本干預相比基於溫度的取樣能產生更高的回應多樣性；其次，在論點生成的案例研究中，STATe的顯性動作序列能捕捉對輸出質量具高度預測性的可解釋特徵；最後，透過分析表現與動作選擇的關聯性，我們能識別動作空間中具潛力卻未被探索的區域，並直接引導生成過程朝向該區域發展。這些成果共同確立STATe作為生成高質量、多樣化且可解釋文本的實用框架。本框架已開源於：https://github.com/zbambergerNLP/state-of-thoughts。

English

Inference-Time-Compute (ITC) methods like Best-of-N and Tree-of-Thoughts are meant to produce output candidates that are both high-quality and diverse, but their use of high-temperature sampling often fails to achieve meaningful output diversity. Moreover, existing ITC methods offer limited control over how to perform reasoning, which in turn limits their explainability. We present STATe-of-Thoughts (STATe), an interpretable ITC method that searches over high-level reasoning patterns. STATe replaces stochastic sampling with discrete and interpretable textual interventions: a controller selects actions encoding high-level reasoning choices, a generator produces reasoning steps conditioned on those choices, and an evaluator scores candidates to guide search. This structured approach yields three main advantages. First, action-guided textual interventions produce greater response diversity than temperature-based sampling. Second, in a case study on argument generation, STATe's explicit action sequences capture interpretable features that are highly predictive of output quality. Third, estimating the association between performance and action choices allows us to identify promising yet unexplored regions of the action space and steer generation directly toward them. Together, these results establish STATe as a practical framework for generating high-quality, diverse, and interpretable text. Our framework is available at https://github.com/zbambergerNLP/state-of-thoughts.