📄 中文摘要
推理是将内部状态与外部输入有意义且语义一致地整合的能力。现代机器学习系统越来越依赖于这种序列推理，涵盖从语言理解到多模态生成的广泛应用，通常在类似于联想记忆模型的原型模式字典上进行操作。理解联想记忆模型中的检索和序列性为深入了解机器学习推理提供了强有力的桥梁。尽管联想记忆模型的静态检索特性已被充分理解，但序列检索和多重记忆整合的理论基础仍然有限，现有研究主要依赖于数值证据。此研究旨在填补这一空白，探讨动态检索的机制及其在复杂推理任务中的应用潜力。
输入驱动的霍普菲尔德网络中的序列检索动态理论
📄 English Summary
A Dynamical Theory of Sequential Retrieval in Input-Driven Hopfield Networks
Reasoning integrates internal states and external inputs in a meaningful and semantically consistent manner. Contemporary machine learning systems increasingly depend on such sequential reasoning, spanning applications from language understanding to multi-modal generation, often operating over dictionaries of prototypical patterns reminiscent of associative memory models. Understanding retrieval and sequentiality in associative memory models provides a powerful bridge to gain insights into ML reasoning. While the static retrieval properties of associative memory models are well understood, the theoretical foundations of sequential retrieval and multi-memory integration remain limited, with existing studies largely relying on numerical evidence. This research aims to address this gap by exploring the mechanisms of dynamic retrieval and its potential applications in complex reasoning tasks.