You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've been tinkering with this repo..
I noticed that whenever I submit any nonsenical, irrelevant input to the chat as a user, the bot pretty much always recognizes it at "greeting" with confidence of over 25%.
For example, as a user in the chat window, I submit "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
The response is "Good to see you again".
I would expect the model to find no valid intents and thus the filter should create an empty list. Instead, the filtered list "results" var always includes the "greeting" result.
What's causing this?
The text was updated successfully, but these errors were encountered:
Hi,
I've been tinkering with this repo..
I noticed that whenever I submit any nonsenical, irrelevant input to the chat as a user, the bot pretty much always recognizes it at "greeting" with confidence of over 25%.
For example, as a user in the chat window, I submit "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
The response is "Good to see you again".
I would expect the model to find no valid intents and thus the filter should create an empty list. Instead, the filtered list "results" var always includes the "greeting" result.
What's causing this?
The text was updated successfully, but these errors were encountered: