The paper refers to studying interactivity models of simultaneous interpreters in the field. It describes the process of simultaneous interpretation as a cognitive event considering two participants in the context of cognitive dissonance and problem solving. The key concept of identifying the models is empathy to a boothmate which has never been studied in the context of Cognitive Translation Theory before. The results of Cognitive Event Analysis show four main interactivity models: empathic productive, empathic unproductive, empathic destructive and nonempathic. The examples of empathic productive interactivity model demonstrate both material, verbal and cognitive empathy. The trajectories of giving a hint show three obligatory cycles of creating productive empathy: identifying cognitive dissonance, illustrating emphatic behavior and inviting to interactivity. The most precious hint happens when deep cognitive empathy between the partners can encourage them by embodied help without any verbal or material hints. It includes two additional cycles: mutual cognitive dissonance and individual attempt to solve the problem. This model is observed mainly in pairs of highly qualified interpreters with enough experience of coupling. Novice simultaneous interpreters are not able to show empathy, so tutors should train the skill of overcoming this gap. Empathic unproductive or destructive interactivity models are observed in pairs of highly qualified interpreters with no experience of coupling because of incongruent mental processes of subject and object.