Hello all, I hope you are doing well.
I am currently working on a project, where a transformer will detect a given URL is a phishing link or a benign link. I created a dataset and fine-tuned it with distilbert-base-uncased; and created this model. But the model throws false positives on links like “open.spotify.com” or “google.com” . I couldn’t figure out what the problem could be. Can anyone help me with this? Thanks so much!