Text this: Comparing Explainable AI Techniques In Language Models: A Case Study For Fake News Detection in Portuguese