Neural Mastery
← Practice
Swish Activation Functioneasy

Swish Activation Function

easyDeep Learning⏱ 10 min
⚡ Libraries allowed · Pure Python earns +10 bonus XP
🎯 Mission
Implement Swish activation f(x) = x * sigmoid(beta * x) widely used in EfficientNet and LLaMA architectures.

Task

Implement `swish(x, beta=1.0)` returning a list of float activation values.

Function Signature

swish(x: list[float], beta: float = 1.0) -> list[float]

Examples

Example 1: Zero & Positive Input
Input: {"x":[0,1,2],"beta":1}
Output: [0,0.7310585786300049,1.7615941559557646]
Example 2: Negative Values
Input: {"x":[-1,-2],"beta":1}
Output: [-0.2689414213699951,-0.23840584404423515]

Constraints

  • Compute element-wise x / (1 + exp(-beta * x)).
Python3Saved ✓
import math

def swish(x, beta=1.0):
"""
Computes Swish activation element-wise: f(x) = x / (1 + exp(-beta * x)).
"""
# Your implementation here
pass

Case 1: Zero & Positive Input
Input: {"x":[0,1,2],"beta":1}
Expected: [0,0.7310585786300049,1.7615941559557646]
Case 2: Negative Values
Input: {"x":[-1,-2],"beta":1}
Expected: [-0.2689414213699951,-0.23840584404423515]