Multimodal large language models excel across diverse domains but struggle with complex visual reasoning tasks. To enhance their reasoning capabilities, current approaches typically rely on explicit search or post-training techniques. However, search-based methods suffer from computational inefficiency due to extensive solution space exploration, while post-training methods demand substantial data, computational resources, and often exhibit training instability. To address these challenges, we propose \textbf{AStar}, a training-free, \textbf{A}utomatic \textbf{S}tructured \textbf{t}hinking paradigm for multimod\textbf{a}l \textbf{r}easoning. Specifically, we introduce novel ``thought cards'', a lightweight library of high-level reasoning patterns abstracted from prior samples. For each test problem, AStar adaptively retrieves the optimal thought cards and seamlessly integrates these external explicit guidelines with the model's internal implicit reasoning capabilities. Compared to previous methods, AStar eliminates computationally expensive explicit search and avoids additional complex post-training processes, enabling a more efficient reasoning approach. Extensive experiments demonstrate that our framework achieves 53.9\% accuracy on MathVerse (surpassing GPT-4o's 50.2\%) and 32.7\% on MathVision (outperforming GPT-4o's 30.4\%). Further analysis reveals the remarkable transferability of our method: thought cards generated from mathematical reasoning can also be applied to other reasoning tasks, even benefiting general visual perception and understanding. AStar serves as a plug-and-play test-time inference method, compatible with other post-training techniques, providing an important complement to existing multimodal reasoning approaches.

翻译：多模态大语言模型在多个领域表现出色，但在复杂的视觉推理任务上仍面临困难。为提升其推理能力，现有方法通常依赖于显式搜索或后训练技术。然而，基于搜索的方法因需探索庞大的解空间而存在计算效率低下的问题，而后训练方法则需要大量数据与计算资源，且常出现训练不稳定的情况。为应对这些挑战，我们提出了 \textbf{AStar}，一种免训练的、用于多模态推理的\textbf{自动化结构化思维范式}。具体而言，我们引入了新颖的“思维卡片”——一种从先前样本中抽象出的高级推理模式的轻量级库。对于每个测试问题，AStar 自适应地检索最优思维卡片，并将这些外部显式指导与模型内部的隐式推理能力无缝集成。与先前方法相比，AStar 消除了计算成本高昂的显式搜索，避免了额外的复杂后训练过程，从而实现了一种更高效的推理途径。大量实验表明，我们的框架在 MathVerse 上达到了 53.9\% 的准确率（超越 GPT-4o 的 50.2\%），在 MathVision 上达到 32.7\%（优于 GPT-4o 的 30.4\%）。进一步分析揭示了我们方法卓越的可迁移性：从数学推理生成的思维卡片同样可应用于其他推理任务，甚至能提升一般的视觉感知与理解能力。AStar 作为一种即插即用的测试时推理方法，可与其他后训练技术兼容，为现有的多模态推理方法提供了重要的补充。