Secrets protected by Qlik Answers guardrails

[taka1]

Hi,
Could you please tell me what are the secrets mentioned in the following document?
I would like to know more with examples.

Here is the description:

https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/QlikAnswers/Qlik-Answers...

Qlik Answers guardrails

Sanitization of Personal Identifying Information (PII) and secrets

Qlik Answers detects and removes PII data and secrets when sending data to the LLM. The detected information is added back when providing responses.

Thank you

[taka1]

It seems that proper nouns (names of people, organizations, places, etc.), dates, time expressions, and pre-defined unique expressions such as quantities, amounts, percentages, etc. could be targeted.
I need to look into the related feature of "Named Entity Extraction in LLM".

Note:

https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/QlikAnswers/Qlik-Answers...

Qlik Answers is offered and supported in the _English language only_, ...