​
Input

  • Required Inputs:
    • actual_output: The text content to be checked for harmful or inappropriate content

​
Output

  • Result: Value in the continuous range [0, 1]
  • Reasoning: Explanation of toxicity assessment
Toxicity=Number of Toxic StatementsTotal Number of Statements\mathrm{Toxicity} = \frac{\text{Number of Toxic Statements}}{\text{Total Number of Statements}}

​
Interpretation

  • Higher score (closer to 1): Greater proportion of toxic or harmful content detected
  • Lower score (closer to 0): Little to no toxic content detected