Role Adherence | Confident AI Docs

Overview

The role adherence metric is a multi-turn metric that uses LLM-as-a-judge to evaluate whether your chatbot consistently maintains its pre-determined role throughout the conversation.

Required Parameters

These are the parameters you must supply in your test case to run evaluations for role adherence metric:

chatbot_role

stringRequired

The role your chatbot has to adhere to throughtout the conversation.

turns

list of TurnRequired

A list of Turns as exchanges between user and assistant.

Parameters of Turn:

role

user | assistantRequired

The role of the person speaking, it’s either user or assistant

content

stringRequired

The content provided by the role for the turn

How Is It Calculated?

The role adherence metric iterates over each assistant turn and uses an LLM to check if the content adheres to the specified chatbot_role.

\text{Role Adherence} = \frac{\text{Number of Assistant Turns that Adhered to Chatbot Role in Conversation}}{\text{Total Number of Assistant Turns in Conversation}}

The final score is the proportion of assisant turns that adhere to the role specified in the conversation.

Create Locally

You can create the RoleAdherenceMetric in deepeval as follows:

1 from deepeval.metrics import RoleAdherenceMetric
2 
3 metric = RoleAdherenceMetric()

Here’s a list of parameters you can configure when creating a RoleAdherenceMetric:

threshold

numberDefaults to 0.5

A float to represent the minimum passing threshold.

model

string | ObjectDefaults to gpt-4.1

A string specifying which of OpenAI’s GPT models to use OR any custom LLM model of type DeepEvalBaseLLM.

include_reason

booleanDefaults to true

A boolean to enable the inclusion a reason for its evaluation score.

async_mode

booleanDefaults to true

A boolean to enable concurrent execution within the measure() method.

strict_mode

booleanDefaults to false

A boolean to enforce a binary metric score: 0 for perfection, 1 otherwise.

verbose_mode

booleanDefaults to false

A boolean to print the intermediate steps used to calculate the metric score.

This can be used for multi-turn E2E

Create Remotely

For users not using deepeval python, or want to run evals remotely on Confident AI, you can use the knowledge retention metric by adding it to a single-turn metric collection. This will allow you to use knowledge retention metric for:

Multi-turn E2E testing
Online and offline evals for traces and spans