# Set up humility rules

> Set up humility rules - Configure humility rules which enable models to recognize, in real-time,
> when they make uncertain predictions or receive data they have not seen before.

This Markdown file sits beside the HTML page at the same path (with a `.md` suffix). It summarizes the topic and lists links for tools and LLM context.

Companion generated at `2026-05-06T18:17:10.037669+00:00` (UTC).

## Primary page

- [Set up humility rules](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/nxt-humility-settings.html): Full documentation for this topic (HTML).

## Sections on this page

- [Create humility rules](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/nxt-humility-settings.html#create-humility-rules): In-page section heading.
- [Manage humility rules](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/nxt-humility-settings.html#manage-humility-rules): In-page section heading.
- [Enable prediction warnings](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/nxt-humility-settings.html#enable-prediction-warnings): In-page section heading.
- [Humility rules considerations](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/nxt-humility-settings.html#humility-rules-considerations): In-page section heading.

## Related documentation

- [NextGen UI documentation](https://docs.datarobot.com/en/docs/workbench/index.html): Linked from this page.
- [Console](https://docs.datarobot.com/en/docs/workbench/nxt-console/index.html): Linked from this page.
- [Deployment settings](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-settings/index.html): Linked from this page.
- [Configure prediction warnings](https://docs.datarobot.com/en/docs/classic-ui/mlops/deployment-settings/humility-settings.html#prediction-warnings): Linked from this page.
- [supports unseen series modeling](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/ts-reference/ts-adv-opt.html#allow-partial-history): Linked from this page.
- [Prediction API documentation](https://docs.datarobot.com/en/docs/api/reference/predapi/legacy-predapi/dr-predapi.html#making-predictions-with-humility-monitoring): Linked from this page.
- [owner](https://docs.datarobot.com/en/docs/reference/misc-ref/roles-permissions.html#deployment-roles): Linked from this page.
- [Predictions Over Time chart](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-monitoring/nxt-data-drift.html#predictions-over-time-chart): Linked from this page.
- [Humilitytab documentation](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-mitigation/nxt-humility.html#enable-prediction-warnings): Linked from this page.

## Documentation content

MLOps allows you to create humility rules for deployments on the Settings > Humility tab. Humility rules enable models to recognize, in real time, when they make uncertain predictions or receive data they have not seen before. Unlike data drift, model humility does not deal with broad statistical properties over time—it is instead triggered for individual predictions, allowing you to set desired behaviors with rules that depend on different triggers. Using humility rules to add triggers and corresponding actions to a prediction helps mitigate risk for models in production. Humility rules help to identify and handle data integrity issues during monitoring and to better identify the root cause of unstable predictions.

The Settings > Humility tab contains the following sub-tabs:

- Settings:Create humility rulesto monitor for uncertainty and specify actions to manage it.
- Prediction Warnings(for regression projects only):Configure prediction warningsto detect when deployments produce predictions with outlier values.

Specific humility rules are available for multiseries projects. While they follow the same general workflow for humility rules as AutoML projects, they have specific settings and options.

## Create humility rules

To create humility rules for a deployment:

1. On theDeploymentsdashboard, open a deployment and navigate to theSettings > Humilitytab.
2. On theHumility Rules Settingspage, if you haven't enabled humility for a model, click theEnable humilitytoggle, then:
3. Click the pencil icon () to enter a name for the rule, then, select aTriggerand anActionto take based on the selected trigger. The trigger detects a rule violation and the action handles the violating prediction. The trigger and action selection process differs for multiseries models: Standard workflowMultiseries workflowSelect aTriggerfor the rule you want to create. Each trigger requires specific settings. The following table and subsequent sections describe these settings. There are three triggers available:TriggerDescriptionTo configureUncertain predictionDetects whether a prediction's value violates the configured thresholds based on lower-bound and upper-bound thresholds for prediction values.Enter values either manually or clickCalculateto use computed thresholds derived from the Holdout partition of the model (DataRobot models only).For regression models, the trigger detects any values outside of the configured thresholds.For binary classification models, the trigger detects any prediction's probability value that isinsidethe thresholds.You can view the type of model for your deployment from the deployment'sOverviewtab.Outlying inputDetects if the input value of a numeric feature is outside of the configured thresholds.Select a numeric feature and set the lower-bound and upper-bound thresholds for its input values. Enter the values manually or clickCalculateto use computed thresholds derived from the training data of the model (DataRobot models only).Low observation regionDetects if the input value of a categorical feature value is not included in the list of specified values.Select a categorical feature and indicate one or more values. Any input value that appears in prediction requests that does not match the indicated values triggers an action.Select anActionfor the rule you are creating. DataRobot applies the action if the trigger indicates a rule violation. There are three actions available:ActionDescriptionTo ConfigureOverride predictionModifies predicted values for rows violating the trigger with the value configured by the action.Set a value that will overwrite the returned value for predictions violating the trigger. For binary classification and multiclass models, the indicated value can be set to either of the model's class labels (e.g., "True" or "False"). For regression models, manually enter a value or use the maximum, minimum, or mean provided by DataRobot (DataRobot models only).Throw errorRows in violation of the trigger return a 480 HTTP error with the predictions, which also contributes to the data error rate on theMonitoring > Service healthtab.Use the default error message provided or specify your own custom error message. This error message will appear along a 480 HTTP error with the predictions.No operationNo changes are made to the detected prediction value.No configuration needed.DataRobot supports multiseries blueprints that support feature derivation and predictions using partial history or no history at all—series that were not trained previously and do not have enough points in the training dataset for accurate predictions. This is useful, for example, in demand forecasting. When a new product is introduced, you may want initial sales predictions. In conjunction with “cold start modeling” (modeling on a series in which there is not sufficient historical data), you can predict on new series, but also keep accurate predictions for serieswitha history. With the support in place, you can set up a humility rule that triggers on a new series (unseen in training data), takes a specified action, and, optionally, returns a custom error message. To do this, take the following steps:Select aTrigger. To include new series data, selectNew seriesas the trigger. This rule detects if a series is present that was not available in the training data and does not have enough history in the prediction data for accurate predictions.Select anAction. Subsequent options are dependent on the selected action, as described in the following table:ActionIf a new series is encountered...Further actionNo operationDataRobot records the event but the prediction is unchanged.N/AUse model with new series supportThe prediction is overridden by the prediction from a selected model with new series support.Select a model thatsupports unseen series modeling. DataRobot preloads supported models in the dropdown.Use global most frequent class (binary classification only)The prediction value is replaced with the most frequent class across all series.N/AUse target mean for all series (regression only)The prediction value is overridden by the global target mean for all series.N/AOverride predictionThe prediction value is changed to the specified preferred value.Enter a numeric value to replace the prediction value for any new series.Return errorThe default or a custom error is returned with the 480 error.Use the default or click in the box to enter a custom error message.If you selectUse model with new series support, when you expand theModel with new series supportdropdown, DataRobot provides a list of models available from Registry, not from Workbench. Using models available from the registry decouples the model from the Use Case and provides support for packages. In this way, you can use a backup model from any compatible Use Case as long as it uses the same target and has the same series available.NoteIf you replace a model within a deployment using a model from a different Use Case, the humility rule is disabled. If the replacement is a model from the same Use Case, the rule is saved. When rule configuration is complete, a rule explanation displays below the rule describing what happens for the configured trigger and respective action.
4. ClickAddto save the rule, and click+Add new ruleto add additional rules.
5. After adding rules, clickSubmit. WarningClickingSubmitis the only way to permanently save new rules and rule changes. If you navigate away from theHumilitytab without clickingSubmit, your rules and edits to rules are not saved. NoteIf a rule is a duplicate of an existing rule, you cannot save it. In this case, when you clickSubmit, a warning displays: After you save and submit the humility rules, DataRobot monitors the deployment using the new rules and any previously created rules. After a rule is created, the prediction response body returns the humility object. Refer to thePrediction API documentationfor more information.

## Manage humility rules

You can edit or delete existing rules from the Humility > Rules tab if you have [owner](https://docs.datarobot.com/en/docs/reference/misc-ref/roles-permissions.html#deployment-roles) permissions:

| Icon | Action | Description |
| --- | --- | --- |
|  | Edit | Change the trigger, action, and associated values for the rule. When finished, click Save Changes. |
|  | Delete | Delete the entire humility rule from the rule list—trigger, action, and values. |
|  | Reorder | Drag and drop the selected humility rule to a new place in the rule list. |

> [!NOTE] Important
> Edits to humility rules can have a significant impact on deployment predictions, as prediction values can be overwritten with new values or can return errors based on the rules configured.

After managing the humility rules, click Submit. If you navigate away from the Humility tab without clicking Submit, your changes will be lost.

**Rule application order**

The displayed list order of your rules determines the order in which they are applied. Although every humility rule trigger is applied, if multiple rules match the trigger of a prediction response, DataRobot applies the first rule in the list that changes the prediction value. However, if any triggered rule has the "Throw Error" action, that rule takes priority.

For example, consider a regression model deployment with the following rules:

| Trigger | Action | Thresholds |
| --- | --- | --- |
| Rule 1: Uncertain Prediction | Override the prediction value to 55. | Lower: 1 Upper: 50 |
| Rule 2: Uncertain Prediction | Override the prediction value to 66. | Lower: 45 Upper: 50 |

If a prediction returns the value 100, both rules will trigger, as both rules detect an uncertain prediction outside of their thresholds. The first rule, Rule 1, takes priority, so the prediction value is overwritten to 55. The action to overwrite the value to 66 (based on Rule 2) is ignored.

In another example, consider a regression model deployment with the following rules:

| Trigger | Action | Thresholds |
| --- | --- | --- |
| Rule 1: Uncertain Prediction | Override the prediction value to 55. | Lower: 1 Upper: 50 |
| Rule 2: Uncertain Prediction | Throw an error. | Lower: 45 Upper: 50 |

If a prediction returns the value 100, both rules will trigger; however, this time, Rule 2 takes priority over Rule 1 because it is configured to return an error. Therefore, the value is not overwritten, as the action to return an error is a higher priority than the numerical order of the rules.

## Enable prediction warnings

Enable prediction warnings for regression model deployments on the Mitigation > Humility > Prediction warnings tab. Prediction warnings allow you to mitigate risk and make models more robust by identifying when predictions do not match their expected result in production. This feature detects when deployments produce predictions with outlier values, summarized in a report that returns with your predictions.

> [!NOTE] Prediction warnings availability
> Prediction warnings are only available for deployments using regression models. This feature does not support classification or time series models.

Prediction warnings provide the same functionality as the Uncertain Prediction trigger that is part of humility monitoring. You may want to enable both, however, because prediction warning results are integrated into the [Predictions Over Time chart](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-monitoring/nxt-data-drift.html#predictions-over-time-chart) on the Data drift tab. For more information, see the [Humilitytab documentation](https://docs.datarobot.com/en/docs/workbench/nxt-console/nxt-mitigation/nxt-humility.html#enable-prediction-warnings).

## Humility rules considerations

Consider the following when using Humility rules:

- You cannot define more than 10 humility rules for a deployment.
- Humility rules can only be defined byownersof the deployment. Users of the deployment can view the rules but cannot edit them or define new rules.
- The "Uncertain Prediction" trigger is only supported for regression and binary classification models.
- Multiclass models only support the "Override prediction" trigger.