提出新方法压缩思维链，保留答案准确率
SCA: Segment-Wise CoT Compression with Answer Alignment
- 按功能分段压缩思维过程，只优化思考部分
- 在多个数据集上压缩率领先，答案准确率不变
- 适合需要高效推理的对话与问答系统
思维链（CoT）推理能提升问题求解能力，但长推理路径会增加推理开销。现有压缩方法通常仅优化完成文本的整体长度。对于结构化思考模型而言，完成文本包含思考段和答案段，因此整体压缩可能同时压缩答案，导致答案漂移。本文提出分段式思维链压缩与答案对齐方法（SCA），通过解析完成文本为功能段，仅将压缩奖励分配给成功思考标记，并通过长度与分布对齐保护答案标记，避免其被压缩。实验表明，在跨多个领域的数据集上，SCA实现了当前最优的思维链压缩效果，同时保持基线模型性能与答案一致性。训练数据与代码已附于补充材料中。
原文摘要 · Abstract (English)
Chain-of-thought (CoT) reasoning improves problem solving, but long think traces increase inference cost. Existing CoT compression methods usually optimize completion-level length. For structured thinking models, however, a completion contains both a think segment and an answer segment, so completion-level compression can save tokens by compressing not only the CoT but also the answer. We call this failure mode answer drift. We propose Segment-wise CoT Compression with Answer Alignment (SCA), an answer-preserving think-compression method. SCA parses completions into functional segments, routes compression rewards only to successful think tokens, and protects answer tokens through length and distribution alignment to a frozen base model. Experiments show that, across datasets from multiple domains, SCA achieves state-of-the-art-level chain-of-thought compression while preserving the base model's performance and answer alignment. Training data and code are included in the supplementary code and data package.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。