The 60-second answer
Activation functions add nonlinearity; without them, stacked linear layers collapse to one linear transformation. ReLU is simple and efficient but can create dead units; GELU is common in Transformers because it provides a smooth gating effect.
Build the answer in this order
1
Give the core idea
Activation functions add nonlinearity; without them, stacked linear layers collapse to one linear transformation.
2
Explain how it works
ReLU is simple and efficient but can create dead units; GELU is common in Transformers because it provides a smooth gating effect.
3
Compare alternatives
Sigmoid and tanh saturate, which can produce small gradients in deep networks.
4
State failure modes + validation
Choose activations together with initialization, normalization, and task architecture.
A useful interview mental model
This is the shape of a strong answer—not a script to memorize.
01Definition
02Mechanism
03Trade-offs
04Failure modes
05When to use
Senior-level signal
- Relate the activation to gradient flow and numerical precision under mixed-precision training.
- In serving, fused activation kernels can materially affect latency at large scale.
What the interviewer is really testing
Mechanistic understanding, assumptions, trade-offs, and whether you can turn a definition into a model decision.
Likely follow-up questions
What assumption makes this approach work?
When would you choose the strongest alternative instead?
What production or data failure mode changes your answer?
Common weak-answer patterns
- Reciting a definition without mechanism or assumptions.
- Claiming one technique is always better without a data regime.
- Stopping before failure modes, validation, or deployment implications.