Response Evaluator
Objective This metric ensures that responses are appropriate and accurate for the given prompt, maintaining consistency.
Interpretation A higher score indicates the response deviates significantly from the prompt or contains inaccuracies. A lower (or zero) score indicates the response is relevant and accurate.
Code Execution
metrics = [
{
"name": "Response Evaluator",
"config": {
"model": "gpt-4o-mini",
"provider": "openai"
},
"column_name": "your-column-identifier",
"schema_mapping": schema_mapping
}
]
Example
Prompt: “Explain Newton’s first law of motion.”
Context: “Newton’s first law states that an object in motion stays in motion unless acted upon by an external force.”
Response: “Newton’s first law is about the gravitational pull of the Earth on objects.”
Metric Output:
{"score": 1, "reason": "Response is inaccurate regarding Newton’s first law."}
Last updated
Was this helpful?