Mindgard home page
Search...
⌘K
Support
Try Mindgard now
Try Mindgard now
Search...
Navigation
Remediations
Preprocess Input Text
Welcome
Introduction
User Guide
Getting Started
Plan AI Risk Management
AI Risk Visibility
Reconnaissance
Customizing your tests
Observations
Policies
AI Risk Remediation
Enterprise Setup
Support & Troubleshooting
Attack Library
Attack Library Overview
Attacks
Remediation Library
Introduction to Remediation
Remediations
ML Output Obfuscation
Apply Context Windows
Separate System Instructions
Dynamically Compile System Prompt
Query Restrictions
Ensemble Methods
Model Hardening
Input Restoration
Homomorphic Encryption
Differential Privacy
Overfitting Detection
Preprocess Input Text
Anonymization of Data
Refine System Prompt
Filter Outputs
Implement Guardrails
On this page
Explanation
How it works
How to implement
Remediations
Preprocess Input Text
Preprocessing prompt inputs.
Preprocessing includes spell check and detection of unknown words, automatic spelling correction, language detection and the use of image processing methods to counter similar-looking characters.
Explanation
How it works
How to implement
Overfitting Detection
Anonymization of Data
Assistant
Responses are generated using AI and may contain mistakes.