Content Moderation
The moderation endpoint serves as a valuable tool for assessing content compliance with QX LABS PTE. LTD.’s usage policies. Developers can utilise this endpoint to identify and address content that may violate our policies, such as filtering it out.
The models categorise content into the following classifications:
Category | Description |
---|---|
Hate | Content that expresses, incites, or promotes hatred on the grounds of race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste. Harassment is extended to include hateful content directed at non-protected groups (e.g., chess players). |
Hate/Threatening | Content that is both hateful and incorporates violence or poses a serious threat to the targeted group based on race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste. |
Harassment | Content that expresses, incites, or promotes the use of harassing language directed at any target. |
Harassment/Threatening | Content involving harassment that also incorporates violence or poses serious harm to any target. |
Self-harm | Content that advocates, fosters, or portrays acts of self-harm, including but not limited to suicide, cutting, and eating disorders. |
Self-harm/Intent | Content in which the speaker communicates their involvement in or intention to participate in acts of self-harm, such as suicide, cutting, and eating disorders. |
Self-harm/Instructions | Content that promotes the engagement in self-harm activities, such as suicide, cutting, and eating disorders, or provides instructions or advice on how to carry out such acts. |
Sexual | Content created with the intent to evoke sexual arousal, including descriptions of sexual activities or the promotion of sexual services (excluding sex education and wellness). |
Sexual/Minors | Content of a sexual nature that involves an individual who is under 18 years old. |
Violence | Content portraying death, violence, or physical harm. |
Violence/Graphic | Content illustrating death, violence, or physical injury with graphic detail. |
The moderation endpoint is free for monitoring QX LABS PTE. LTD. API inputs and outputs, with other use cases currently disallowed. Note that accuracy may decrease for longer text pieces. For improved accuracy, consider breaking down lengthy content into smaller chunks, each less than 2,000 characters.