This repository contains explainability and interpretability tools for the mosquito classification CNN model trained in the Mosaico_train repository.

• GradCAM
• GradCAM++
• Guided GradCAM
• Other saliency visualization techniques

These methods help verify that the model bases its predictions on relevant visual features, consistent with expert entomologist knowledge.

• Provide visual explanations of model decisions.
• Support model validation and debugging.
• Increase trust in model predictions through transparency.