As models constitute the main software artifacts in model-driven engineering, managing their evolution has attracted much research. One important technique in this realm is model differencing concerned with identifying differences among models. The majority of existing model differencing approaches identify differences by applying a fine-grained analysis of models based on their abstract syntax representation. Thereby, syntactic differences among models can be revealed. However, syntactic differences can only approximate semantic differences among models. The authors propose a generic framework for realizing semantic model differencing operators revealing semantic differences among models.