Read news on wrong answers with our app.
Read more in the app
Method prevents an AI model from being overconfident about wrong answers