Struct HallucinationEvaluator 

pub struct HallucinationEvaluator { /* private fields */ }

Evaluates whether agent responses are grounded (not hallucinated).

Uses an LLM-as-judge to assess whether the model’s claims are supported by the conversation context, tool outputs, and provided information.

Implementations

impl HallucinationEvaluator

pub fn new() -> Self

Create a new hallucination evaluator.

pub fn with_intermediate(self, eval: bool) -> Self

Set whether to evaluate intermediate responses in addition to the final response.

pub fn with_judge_model(self, model: impl Into<String>) -> Self

Set an override judge model name.

pub fn with_llm(self, llm: Arc<dyn BaseLlm>) -> Self

Provide an LLM instance for performing evaluations.

Trait Implementations

impl Default for HallucinationEvaluator

fn default() -> Self

Returns the “default value” for a type.

impl Evaluator for HallucinationEvaluator

fn evaluate<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, actual: &'life1 [Invocation], _expected: Option<&'life2 [Invocation]>, ) -> Pin<Box<dyn Future<Output = Result<EvalResult, EvalError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Evaluate agent invocations.