GenerateLossClustersResponse

Response message for EvaluationAnalyticsService.GenerateLossClusters.

Fields
analysisTime string (Timestamp format)

The timestamp when this analysis was completed.

Uses RFC 3339, where generated output will always be Z-normalized and use 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: "2014-10-02T15:01:23Z", "2014-10-02T15:01:23.045123456Z" or "2014-10-02T15:01:23+05:30".

results[] object (LossAnalysisResult)

The analysis results, one per config provided in the request.

JSON representation 
{
  "analysisTime": string,
  "results": [
    {
      object (LossAnalysisResult)
    }
  ]
}

LossAnalysisResult

The top-level result for loss analysis, stored within an EvalSet.

Fields
config object (LossAnalysisConfig)

The configuration used to generate this analysis.

analysisTime string (Timestamp format)

The timestamp when this analysis was performed.

Uses RFC 3339, where generated output will always be Z-normalized and use 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: "2014-10-02T15:01:23Z", "2014-10-02T15:01:23.045123456Z" or "2014-10-02T15:01:23+05:30".

clusters[] object (LossCluster)

The list of identified loss clusters.

JSON representation 
{
  "config": {
    object (LossAnalysisConfig)
  },
  "analysisTime": string,
  "clusters": [
    {
      object (LossCluster)
    }
  ]
}

LossCluster

Represents a semantic grouping of failures (e.g., "Hallucination of Action").

Fields
clusterId string

Unique identifier for the loss cluster within the scope of the analysis result.

taxonomyEntry object (LossTaxonomyEntry)

The structured definition of the loss taxonomy for this cluster.

itemCount integer

The total number of EvaluationItems falling into this cluster.

examples[] object (LossExample)

A list of examples that belong to this cluster. This links the cluster back to the specific EvaluationItems and Rubrics.

JSON representation 
{
  "clusterId": string,
  "taxonomyEntry": {
    object (LossTaxonomyEntry)
  },
  "itemCount": integer,
  "examples": [
    {
      object (LossExample)
    }
  ]
}

LossTaxonomyEntry

Defines a specific entry in the loss pattern taxonomy.

Fields
l1Category string

The primary category of the loss (e.g., "Hallucination", "Tool Calling"). This field is typically required.

l2Category string

The secondary category of the loss (e.g., "Hallucination of Action", "Incorrect Tool Selection").

description string

A detailed description of this loss pattern. Example: "The agent verbally confirms an action without executing the tool."

JSON representation 
{
  "l1Category": string,
  "l2Category": string,
  "description": string
}

LossExample

Represents a specific example of a loss pattern.

Fields
failedRubrics[] object (FailedRubric)

The specific rubric(s) that failed and caused this example to be classified here. An example might fail multiple rubrics, but only specific ones trigger this loss pattern.

source Union type
The source of this example. source can be only one of the following:
evaluationItem string

Reference to the persisted EvalItem resource name. Format: projects/.../locations/.../evaluationItems/{itemId} Used when analysis is run on an EvalSet.

JSON representation 
{
  "failedRubrics": [
    {
      object (FailedRubric)
    }
  ],

  // source
  "evaluationItem": string
  // Union type
}