连接效率与透明性：可解释的CoT压缩在多模态大推理模型中
Bridging Efficiency and Transparency: Explainable CoT Compression in Multimodal Large Reasoning Models
- School of Computer Science and Engineering, Southeast University（东南大学计算机科学与工程学院）
- Key Laboratory of Computer Network and Information Integration (SEU), Ministry of Education, China（计算机网络与信息集成重点实验室（SEU））
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
XMCC通过强化学习优化的顺序决策过程，实现多模态大推理模型中可解释的CoT压缩，同时保持推理正确性和生成可解释的压缩解释。
AI中文摘要：
长链推理（Long CoTs）在多模态推理模型中被广泛用于解决复杂任务，通过捕捉详细的视觉信息。然而，这些长CoTs通常过于冗长，包含冗余的推理步骤，这会阻碍推理效率。压缩这些长CoTs是一种自然的解决方案，但现有方法面临两个主要挑战：（1）它们可能通过移除必要的对齐提示而损害视觉-文本推理的完整性；（2）压缩过程缺乏可解释性，使得难以确定哪些信息是关键的。为了解决这些问题，我们提出了XMCC，一种可解释的多模态CoT压缩器，将压缩过程建模为一个通过强化学习优化的顺序决策过程。XMCC可以有效缩短推理轨迹，同时保留关键推理步骤和答案的正确性，并同时生成自然语言的解释，以说明其压缩决策。在具有代表性的多模态推理基准上的广泛实验表明，XMCC不仅减少了推理长度，还提供了可解释的解释，验证了其有效性。
英文摘要：
Long chains of thought (Long CoTs) are widely employed in multimodal reasoning models to tackle complex tasks by capturing detailed visual information. However, these Long CoTs are often excessively lengthy and contain redundant reasoning steps, which can hinder inference efficiency. Compressing these long CoTs is a natural solution, yet existing approaches face two major challenges: (1) they may compromise the integrity of visual-textual reasoning by removing essential alignment cues, and (2) the compression process lacks explainability, making it difficult to discern which information is critical. To address these problems, we propose XMCC, an eXplainable Multimodal CoT Compressor that formulates compression as a sequential decision-making process optimized via reinforcement learning. XMCC can effectively shorten reasoning trajectories while preserving key reasoning steps and answer correctness, and simultaneously generates natural-language explanations for its compression decisions. Extensive experiments on representative multimodal reasoning benchmarks demonstrate that XMCC not only reduces reasoning length but also provides explainable explanations, validating its effectiveness.
