> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mindgard.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Model Hardening

> Make AI models more robust to adversarial inputs.

export const what_0 = "Use adversarial robustness techniques to re-train or re-deploy your model securely."

export const how_0 = "Reduces the success of the attacker by making the model more resilient to attacks."

export const implement_0 = "Adversarial training, network distillation."

This technique strives to make AI models more robust to adversarial inputs via adversarial training or network distillation. Examples include (i) using randomization to inject noise during training to enhance resilience to evasion attacks (especially triggered by subtle perturbations), (ii) Gradient Masking, (iii) Feature Squeezing.

## Explanation

{what_0} 

## How it works

{how_0} 

## How to implement

{implement_0}