AI 模型中稀疏符号概念出现证明的进展（英文版）

英文标题：Where We Have Arrived in Proving the Emergence of Sparse Symbolic Concepts in AI Models中文摘要：通过证明 AI 模型对不同交互概念的输入输出方式的映射，本研究试图证明，大多数 AI 模型可使用少数交互概念来在任意遮掩样本上模拟模型输出。英文摘要：This paper aims to prove the emergence of symbolic concepts in well-trainedAI models. We prove that if (1) the high-order derivatives of the model outputw.r.t. the input variables are all zero, (2) the AI model can be used onoccluded samples and will yield higher confidence when the input sample is lessoccluded, and (3) the c