Parameters
Persona description for the simulated user.
The goal or intent the persona is pursuing. Use this to specify what the simulater user is trying to achieve
during the interaction. For example,
"To ask questions about world history".Input passed to the persona. In multi-turn conversations, you can connect an Agent node’s
output to a Persona node’s input to simulate back-and-forth dialogue. In this case, this parameter can be
set to
{input} to use the node’s input or {@Agent.output} to explicitly reference the upstream Agent node’s output
(assuming the Agent node’s ID is Agent).Initial message from the persona. If set, this node will always emit this message on the first turn of the interaction.
Reference trajectory to replay or condition the persona. You can simply select the
Benchmark Trigger
node that supplies this trajectory.ID of the node persona should mimic from the reference trajectory.
Maximum number of turns to simulate.
LLM used to simulate the persona.
Credentials
Select an LLM Credential used by the persona model.
Inputs
Input passed into the persona.
Outputs
Persona response message.