Hyperbolic Tangent (tanh)
Definition
tanh smoothly maps numbers to a range between minus one and one.
The hyperbolic tangent is a smooth function that maps real inputs to values between minus one and one and is used as an activation function.
Simply explained
Like a soft limiter: even very large or very small inputs are compressed to values close to plus one or minus one, respectively.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A recurrent network limits candidate states with tanh to a symmetric value range.
Limits and pitfalls
- For large values, the function saturates and gradients become small; it is not the best choice for every layer.
Origin & history
Tanh was widespread in early neural networks and remains relevant in recurrent architectures and mathematical models.
Use in mid-sized companies
- Possible application: A recurrent network limits candidate states with tanh to a symmetric value range.
Frequently asked questions
What does Hyperbolic Tangent (tanh) mean?
The hyperbolic tangent is a smooth function that maps real inputs to values between minus one and one and is used as an activation function. What should you watch out for? For large values, the function saturates and gradients become small; it is not the best choice for every layer.
Sources
- Efficient BackProp · LeCun et al., Efficient BackProp
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: