菜单

关于 🐙 GitHub
arXiv 提交日期: 2025-12-14
📄 Abstract - State over Tokens: Characterizing the Role of Reasoning Tokens

Large Language Models (LLMs) can generate reasoning tokens before their final answer to boost performance on complex tasks. While these sequences seem like human thought processes, empirical evidence reveals that they are not a faithful explanation of the model's actual reasoning process. To address this gap between appearance and function, we introduce the State over Tokens (SoT) conceptual framework. SoT reframes reasoning tokens not as a linguistic narrative, but as an externalized computational state -- the sole persistent information carrier across the model's stateless generation cycles. This explains how the tokens can drive correct reasoning without being a faithful explanation when read as text and surfaces previously overlooked research questions on these tokens. We argue that to truly understand the process that LLMs do, research must move beyond reading the reasoning tokens as text and focus on decoding them as state.

顶级标签: llm theory model evaluation
详细标签: reasoning tokens chain-of-thought interpretability faithfulness state representation 或 搜索:

状态优于标记:重新概念化大语言模型中的推理标记 / State over Tokens: Characterizing the Role of Reasoning Tokens


1️⃣ 一句话总结

本文提出了“状态优于标记”的概念框架,认为大语言模型在最终答案前生成的推理标记序列,本质上是外部化的计算状态载体,而非对人类思维过程的忠实解释性文本。


2️⃣ 论文创新点

1. State over Tokens (SoT) 框架

2. 推理标记与推理文本的区分

3. 揭示编码任意性与共享意义误解

4. 提出文本与状态的本体论分歧

5. 审视语言作为计算媒介的特殊性

6. 揭示SoT作为忠实解释的根本张力


3️⃣ 主要结果与价值

结果亮点

实际价值


4️⃣ 术语表

源自 arXiv: 2512.12777