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
FallbackPolicy or the
TwoStageFallbackPolicy to your
FallbackPolicy has one fallback action, which will
be executed if the intent recognition has a confidence below
or if none of the dialogue policies predict an action with
confidence higher than
The thresholds and fallback action can be adjusted in the policy configuration
file as parameters of the
policies: - 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
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:
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
your policy configuration file. For example:
policies: - 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
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
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¶
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
If a NLU prediction has a low confidence score, the user is asked to affirm the classification of the intent. (Default action:
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:
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:
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_nameis triggered (e.g. a handoff to a human).
Rasa Core provides the default implementations of
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.
policies: - 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"