The 60-second answer
L1 penalizes absolute parameter magnitude and can drive coefficients exactly to zero; L2 penalizes squared magnitude and tends to shrink parameters smoothly. L1 can act as feature selection in sparse linear settings, while L2 is stable when correlated features should share weight.
Build the answer in this order
L1 penalizes absolute parameter magnitude and can drive coefficients exactly to zero; L2 penalizes squared magnitude and tends to shrink parameters smoothly.
L1 can act as feature selection in sparse linear settings, while L2 is stable when correlated features should share weight.
Standardize features before interpreting regularization strength in linear models.
Choose the coefficient through validation and discuss elastic net when both sparsity and stability are useful.
A useful interview mental model
This is the shape of a strong answer—not a script to memorize.
Senior-level signal
- Senior answers note that neural-network “weight decay” is not always mathematically identical to adding L2 under adaptive optimizers.
- Explain how correlated features can make sparse feature-selection interpretations unstable.
What the interviewer is really testing
Likely follow-up questions
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.