📄 Manipulative Email¶

metrics.llm_as_judge.direct.criteria.manipulative_email

CriteriaWithOptions(
    name="manipulative_email",
    description="Does the email response sound manipulative?",
    prediction_field="email response",
    context_fields=[],
    options=[
        CriteriaOption(
            name="Yes",
            description="The email in the response is written in an exaggerated way, it is subjective, and trying to convince readers to buy a product they may not really want.",
        ),
        CriteriaOption(
            name="No",
            description="The email in the response is objectively highlighting features of a product without exaggeration or trying to manipulate the reader into buying this product.",
        ),
    ],
    option_map={
        "Yes": 1.0,
        "No": 0.0,
    },
)

[source]

from unitxt.llm_as_judge_constants import CriteriaOption

Explanation about CriteriaWithOptions¶

Criteria used by DirectLLMJudge to run evaluations.

Explanation about CriteriaOption¶

A criteria option.