# NLU predicts nlu\_fallback even the confidence score of some other intent is higher than the threshhold

**URL:** <https://forum.rasa.com/t/nlu-predicts-nlu-fallback-even-the-confidence-score-of-some-other-intent-is-higher-than-the-threshhold/48929>\
**Category:** Rasa Open Source\
**Created:** [November 2, 2021, 6:56am UTC](https://forum.rasa.com/t/nlu-predicts-nlu-fallback-even-the-confidence-score-of-some-other-intent-is-higher-than-the-threshhold/48929 "2021-11-02T06:56:30Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![wolfias](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/wolfias/32/15488_2.png) [@wolfias](https://forum.rasa.com/u/wolfias)\
**Post date:** [November 2, 2021, 6:56am UTC](https://forum.rasa.com/t/nlu-predicts-nlu-fallback-even-the-confidence-score-of-some-other-intent-is-higher-than-the-threshhold/48929/1 "2021-11-02T06:56:30Z")

</div>

Hi, when I added the nlu\_fallback to my rasa config, it didn’t worked as expectedly. NLU predicts nlu\_fallback even when the confidence score of some other intent is higher than the threshhold. Can anyone help? Thanks!

the 2 test case as follows:

1. the nlu result of the config without nlu\_fallback: ![njUdY0dnby](https://europe1.discourse-cdn.com/flex013/uploads/rasa/original/3X/f/3/f3f1479121a52c3afe2361b92556a9688130de85.png)

2. the nlu result of the config with nlu\_fallback (threshold=0.7): ![4LE3a1tQXP](https://europe1.discourse-cdn.com/flex013/uploads/rasa/original/3X/2/6/26222480395fd51e1bc2520ef5ac6219beaf35c4.png)

the rasa version: 2.7.1

my rasa config:

```auto
language: zh
pipeline:
#- name: WhitespaceTokenizer
  - name: JiebaTokenizer
    intent_tokenization_flag: True
    intent_split_symbol: "+"
  - name: RegexFeaturizer
  - name: LexicalSyntacticFeaturizer
  - name: CountVectorsFeaturizer
  - name: CountVectorsFeaturizer
    analyzer: char_wb
    min_ngram: 1
    max_ngram: 4
  - name: DIETClassifier
    epochs: 100
    model_confidence: softmax
    constrain_similarities: true
  # - name: FallbackClassifier
  # threshold: 0.7
  - name: EntitySynonymMapper
  - name: ResponseSelector
    epochs: 100
    constrain_similarities: true
  - name: ResponseSelector
    epochs: 100
    retrieval_intent: faq
  - name: ResponseSelector
    epochs: 100
    retrieval_intent: chitchat
# Configuration for Rasa Core.
# https://rasa.com/docs/rasa/core/policies/
policies:
  - name: MemoizationPolicy
  - name: TEDPolicy
    max_history: 5
    epochs: 100
    constrain_similarities: true
  - name: RulePolicy

```

---

<div class="post-metadata">

**Author:** ![wolfias](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/wolfias/32/15488_2.png) [@wolfias](https://forum.rasa.com/u/wolfias)\
**Post date:** [November 3, 2021, 7:34am UTC](https://forum.rasa.com/t/nlu-predicts-nlu-fallback-even-the-confidence-score-of-some-other-intent-is-higher-than-the-threshhold/48929/2 "2021-11-03T07:34:08Z")

</div>

Is there anyone meeting this problem? All help and suggestions would really be appreiciated.
