AI engineering glossary · Calculus optimisation
What is SAM (sharpness-aware minimisation)?
Sharpness-aware minimisation (also called SAM): A training method that seeks parameters lying in neighbourhoods with uniformly low loss. Each step first perturbs the weights in the direction that raises the loss most, then takes the gradient there, which steers optimisation toward flat minima.
Sharpness-aware minimisation: an example
Each SAM update needs two gradient computations, one at the current weights and one at the perturbed weights, roughly doubling the cost per step.
Related terms
More Calculus optimisation terms
- Simulated annealing
- Stationary point
- Step decay
- Stochastic gradient descent
- Strong convexity
- Subgradient
- Subgradient method
- Sum rule
Where you learn it
Sharpness-aware minimisation is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.