What is broadcasting in PyTorch?

Updated May 17, 2026

Short answer

Broadcasting allows operations on tensors of different shapes.

Deep explanation

PyTorch automatically expands smaller tensors to match larger ones without copying data.

Real-world example

Used in vectorized operations in deep learning.

Common mistakes

  • Assuming actual memory expansion occurs.

Follow-up questions

  • What are broadcasting rules?
  • When does broadcasting fail?

More PyTorch interview questions

View all →