Accordion-Thinking：通过动态摘要实现的自我调节步骤摘要以提高LLM推理的效率和可读性
Accordion-Thinking: Self-Regulated Step Summaries for Efficient and Readable LLM Reasoning
AI总结：
本文提出Accordion-Thinking框架，通过动态摘要实现LLM推理的自我调节，减少历史token依赖，提升效率和可读性，实验显示其在48GB GPU内存配置下实现三倍吞吐量且保持准确率。
摘要：
通过动态摘要实现的自我调节步骤摘要，使LLM推理更高效且可读。本文提出Accordion-Thinking框架，使LLM学会通过动态摘要自我调节推理步骤的粒度。该机制启用折叠推理模式，模型定期总结其思考过程并丢弃先前思考以减少对历史token的依赖。我们应用强化学习进一步激励这一能力，发现一个关键见解：折叠模式与彻底展开模式之间的准确性差距逐渐缩小并在训练过程中最终消失。这表明模型学会将必要的推理信息编码到紧凑的摘要中，实现了推理上下文的有效压缩。我们的Accordion-Thinking表明，通过学习自我压缩，LLM能够以最小的依赖token开销处理复杂推理任务而不影响解决方案质量，并在48GB GPU内存配置下实现三倍吞吐量且保持准确率，而结构化的步骤摘要提供了人类可读的推理过程记录。
英文摘要：
Scaling test-time compute via long Chain-of-Thought unlocks remarkable gains in reasoning capabilities, yet it faces practical limits due to the linear growth of KV cache and quadratic attention complexity. In this paper, we introduce Accordion-Thinking, an end-to-end framework where LLMs learn to self-regulate the granularity of the reasoning steps through dynamic summarization. This mechanism enables a Fold inference mode, where the model periodically summarizes its thought process and discards former thoughts to reduce dependency on historical tokens. We apply reinforcement learning to incentivize this capability further, uncovering a critical insight: the accuracy gap between the highly efficient Fold mode and the exhaustive Unfold mode progressively narrows and eventually vanishes over the course of training. This phenomenon demonstrates that the model learns to encode essential reasoning information into compact summaries, achieving effective compression of the reasoning context. Our Accordion-Thinking demonstrates that with learned self-compression, LLMs can tackle complex reasoning tasks with minimal dependency token overhead without compromising solution quality, and it achieves a three times throughput while maintaining accuracy on a 48GB GPU memory configuration, while the structured step summaries provide a human-readable account of the reasoning process.
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。
