Output only. feature attributions grouped by predicted outputs.
For Models that predict only one output, such as regression Models that predict only one score, there is only one attibution that explains the predicted output. For Models that predict multiple outputs, such as multiclass Models that predict multiple classes, each element explains one specific item. Attribution.output_index can be used to identify which output this attribution is explaining.
By default, we provide Shapley values for the predicted class. However, you can configure the explanation request to generate Shapley values for any other classes too. For example, if a model predicts a probability of 0.4 for approving a loan application, the model's decision is to reject the application since p(reject) = 0.6 > p(approve) = 0.4, and the default Shapley values would be computed for rejection decision and not approval, even though the latter might be the positive class.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-06-27 UTC."],[],[],null,["# Explanation of a prediction (provided in [PredictResponse.predictions](/vertex-ai/docs/reference/rest/v1beta1/PredictResponse#FIELDS.predictions)) produced by the Model on a given [instance](/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints/explain#body.request_body.FIELDS.instances).\nFields `attributions[]` `object (`[Attribution](/vertex-ai/docs/reference/rest/v1beta1/ModelExplanation#Attribution)`)` \nOutput only. feature attributions grouped by predicted outputs.\n\nFor Models that predict only one output, such as regression Models that predict only one score, there is only one attibution that explains the predicted output. For Models that predict multiple outputs, such as multiclass Models that predict multiple classes, each element explains one specific item. [Attribution.output_index](/vertex-ai/docs/reference/rest/v1beta1/ModelExplanation#Attribution.FIELDS.output_index) can be used to identify which output this attribution is explaining.\n\nBy default, we provide Shapley values for the predicted class. However, you can configure the explanation request to generate Shapley values for any other classes too. For example, if a model predicts a probability of `0.4` for approving a loan application, the model's decision is to reject the application since `p(reject) = 0.6 \u003e p(approve) = 0.4`, and the default Shapley values would be computed for rejection decision and not approval, even though the latter might be the positive class.\n\nIf users set [ExplanationParameters.top_k](/vertex-ai/docs/reference/rest/v1beta1/ExplanationSpec#ExplanationParameters.FIELDS.top_k), the attributions are sorted by [instanceOutputValue](/vertex-ai/docs/reference/rest/v1beta1/ModelExplanation#Attribution.FIELDS.instance_output_value) in descending order. If [ExplanationParameters.output_indices](/vertex-ai/docs/reference/rest/v1beta1/ExplanationSpec#ExplanationParameters.FIELDS.output_indices) is specified, the attributions are stored by [Attribution.output_index](/vertex-ai/docs/reference/rest/v1beta1/ModelExplanation#Attribution.FIELDS.output_index) in the same order as they appear in the outputIndices.\n`neighbors[]` `object (`[Neighbor](/vertex-ai/docs/reference/rest/v1beta1/Explanation#Neighbor)`)` \nOutput only. List of the nearest neighbors for example-based explanations.\n\nFor models deployed with the examples explanations feature enabled, the attributions field is empty and instead the neighbors field is populated. \n\nNeighbor\n--------\n\nNeighbors for example-based explanations.\nFields `neighborId` `string` \nOutput only. The neighbor id.\n`neighborDistance` `number` \nOutput only. The neighbor distance."]]