Class SafetyFilterOptions<T>

Namespace

AiDotNet.Models.Options

Assembly

AiDotNet.dll

Configuration options for safety filtering mechanisms.

public class SafetyFilterOptions<T>

Type Parameters

T

The numeric data type used for calculations (e.g., float, double).

Inheritance

object

SafetyFilterOptions<T>

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

These options control how inputs and outputs are validated and filtered to prevent harmful or inappropriate content from passing through the AI system.

For Beginners: These settings control how strict your "security guards" are. You can adjust sensitivity thresholds, what types of content to filter, and how thoroughly to check for problems.

Properties

EnableInputValidation

Gets or sets whether to enable input validation.

public bool EnableInputValidation { get; set; }

Property Value

bool

True to validate inputs, false otherwise (default: true).

Remarks

For Beginners: Input validation checks requests before processing them, catching malicious or malformed inputs early.

EnableOutputFiltering

Gets or sets whether to enable output filtering.

public bool EnableOutputFiltering { get; set; }

Property Value

bool

True to filter outputs, false otherwise (default: true).

Remarks

For Beginners: Output filtering checks responses before showing them to users, ensuring no harmful content gets through.

HarmfulContentCategories

Gets or sets the harmful content categories to check for.

public string[] HarmfulContentCategories { get; set; }

Property Value

string[]

Array of category names, defaulting to common harmful categories.

Remarks

For Beginners: These are the types of harmful content to watch for, like violence, hate speech, adult content, etc.

JailbreakSensitivity

Gets or sets the jailbreak detection sensitivity.

public double JailbreakSensitivity { get; set; }

Property Value

double

The sensitivity (0-1), defaulting to 0.7.

Remarks

For Beginners: How sensitive the system is to jailbreak attempts. Higher values catch more jailbreaks but might have false positives.

LogFilePath

Gets or sets the file path used when logging filtered content.

public string? LogFilePath { get; set; }

Property Value

string

Remarks

When null or empty, a default relative path is used. Prefer an absolute path in production deployments or integrate with a logging framework.

LogFilteredContent

Gets or sets whether to log filtered content for review.

public bool LogFilteredContent { get; set; }

Property Value

bool

True to log filtered content, false otherwise (default: true).

Remarks

For Beginners: Logging helps you review what's being filtered and improve the filtering over time.

MaxInputLength

Gets or sets the maximum input length to process.

public int MaxInputLength { get; set; }

Property Value

int

The maximum length, defaulting to 10000.

Remarks

For Beginners: Inputs longer than this are rejected to prevent abuse and ensure processing efficiency.

SafetyThreshold

Gets or sets the safety threshold for content filtering.

public double SafetyThreshold { get; set; }

Property Value

double

The threshold (0-1), defaulting to 0.8.

Remarks

For Beginners: Content with safety scores below this threshold is flagged. Higher thresholds (closer to 1) are more strict and filter more content.

UseClassifier

Gets or sets whether to use a classifier for harmful content detection.

public bool UseClassifier { get; set; }

Property Value

bool

True to use classifier, false for rule-based detection (default: true).

Remarks

For Beginners: Classifiers use machine learning to detect harmful content and are generally more accurate than simple rules.