Ban List

Block specific banned terms or patterns. Use guardrails to control AI output safety and compliance.

Objective This metric filters out responses containing banned words or phrases, ensuring compliance with content guidelines.

Interpretation A higher score indicates the response contains banned terms. A lower (or zero) score indicates no banned terms were found.

Code Execution

metrics = [
    {
        "name": "Ban List",
        "config": {
            "model": "gpt-4o-mini",
            "provider": "openai"
        },
        "column_name": "your-column-identifier",
        "schema_mapping": schema_mapping
    }
]

Example

• Prompt: “Tell me a story.”

• Context: “Avoid the use of these banned words: [redacted list].”

• Response: “Once upon a time, [banned word]…”

• Metric Output: {"score": 1, "reason": "Response contains a banned word or phrase."}