The field of artificial intelligence is moving towards developing more transparent and explainable models, particularly in image classification tasks. Recent research has focused on creating models that can provide insights into their decision-making processes, making them more trustworthy and reliable. One of the key directions is the development of concept-based explanations, which decompose AI predictions into human-understandable concepts. Another important area is the creation of models that can generate minimal and faithful explanations for their decisions, making it easier to identify the most critical regions of an image that contribute to the prediction. Furthermore, researchers are working on improving the efficiency and effectiveness of counterfactual explanations, which can help identify the smallest changes to input features needed to change the model's prediction. Noteworthy papers include: CE-FAM, which proposes a novel concept-based explanation method that employs a branched network to learn concepts and their corresponding regions. ACE, which introduces a sample-efficient algorithm for generating counterfactual explanations. TextCAM, which enriches Class Activation Mapping with natural languages to provide more semantic insights into the model's decisions.