Advanced Deep Learning Interview Questions, issue 22, Apr 12, 2026
The Perfect Discriminator Trap
Senior Machine Learning Engineer interview at Google DeepMind, and the interviewer asks:
“Your newly initialized standard GAN is suffering from vanishing gradients right out of the gate, and the discriminator’s accuracy is sitting at exactly 100%. How do you fix it?”
A 100% accurate discriminator isn’t “too strong”, it’s a symptom that your loss function has collapsed into a zero-gradient regime.
The full answer, with the mechanism and the arithmetic, is for paid subscribers on Substack.