Fallback Actions

Sometimes you want to revert to a fallback action, such as replying, “Sorry, I didn’t understand that”. You can handle fallback cases by adding either the FallbackPolicy or the TwoStageFallbackPolicy to your policy ensemble.

Fallback Policy

The FallbackPolicy has one fallback action, which will be executed if the intent recognition has a confidence below nlu_threshold or if none of the dialogue policies predict an action with confidence higher than core_threshold.

The thresholds and fallback action can be adjusted in the policy configuration file as parameters of the FallbackPolicy.

  - name: "FallbackPolicy"
    nlu_threshold: 0.4
    core_threshold: 0.3
    fallback_action_name: "action_default_fallback"

action_default_fallback is a default action in Rasa Core which sends the utter_default template message to the user. Make sure to specify the utter_default in your domain file. It will also revert back to the state of the conversation before the user message that caused the fallback, so that it will not influence the prediction of future actions. You can take a look at the source of the action below:

class rasa.core.actions.action.ActionDefaultFallback

Executes the fallback action and goes back to the previous state of the dialogue

You can also create your own custom action to use as a fallback (see custom actions for more info on custom actions). If you do, make sure to pass the custom fallback action to FallbackPolicy inside your policy configuration file. For example:

  - name: "FallbackPolicy"
    nlu_threshold: 0.4
    core_threshold: 0.3
    fallback_action_name: "my_fallback_action"


If your custom fallback action does not return a UserUtteranceReverted event, the next predictions of your bot may become inaccurate, as it is very likely that the fallback action is not present in your stories.

If you have a specific intent, let’s say it’s called out_of_scope, that should always trigger the fallback action, you should add this as a story:

## fallback story
* out_of_scope
  - action_default_fallback

Two-stage Fallback Policy

The TwoStageFallbackPolicy handles low NLU confidence in multiple stages by trying to disambiguate the user input (low core confidence is handled in the same manner as the FallbackPolicy).

  • If a NLU prediction has a low confidence score, the user is asked to affirm the classification of the intent. (Default action: action_default_ask_affirmation)

    • If they affirm, the story continues as if the intent was classified with high confidence from the beginning.
    • If they deny, the user is asked to rephrase their message.
  • Rephrasing (default action: action_default_ask_rephrase)

    • If the classification of the rephrased intent was confident, the story continues as if the user had this intent from the beginning.
    • If the rephrased intent was not classified with high confidence, the user is asked to affirm the classified intent.
  • Second affirmation (default action: action_default_ask_affirmation)

    • If the user affirms the intent, the story continues as if the user had this intent from the beginning.
    • If the user denies, the original intent is classified as the specified deny_suggestion_intent_name, and an ultimate fallback action fallback_nlu_action_name is triggered (e.g. a handoff to a human).

Rasa Core provides the default implementations of action_default_ask_affirmation and action_default_ask_rephrase. The default implementation of action_default_ask_rephrase action utters the response template utter_ask_rephrase, so be sure to specify this template in your domain file. The implementation of both actions can be overwritten with custom actions.

You can specify the core fallback action as well as the ultimate NLU fallback action as parameters to TwoStageFallbackPolicy in your policy configuration file.

  - name: TwoStageFallbackPolicy
    nlu_threshold: 0.3
    core_threshold: 0.3
    fallback_core_action_name: "action_default_fallback"
    fallback_nlu_action_name: "action_default_fallback"
    deny_suggestion_intent_name: "out_of_scope"