Skip to main content
T3 Chips and hardware Houmo AI Internship

Houmo AI Infra internship

Houmo AI Infra internship interview questions covering inference optimization

Inference optimization

Fundamentals

  1. What is the basic idea of gradient descent? How do you differentiate with the chain rule?
  2. How is the derivative of ReLU defined at zero?
  3. How is Softmax implemented? How do you handle numerical overflow?

Inference optimization

  1. Explain the core idea of AdaQuant.
  2. Explain the key technical points of I-ViT quantization.