Q1 How we used datasets Operation of FFNN Hyper-parameter settings Learning curves (training and test perplexity) Fakes Detection results (confusion matrix) Limitations of our approach and possible solutions Q2 How we used datasets Operation of LSTM LLM Hyper-parameter settings Learning curves (training and test perplexity) Fakes Detection results (confusion matrix) Limitations of our approach and possible solutions Q3 How you selected the best model Limitations of our approach and possible solutions