This topic describes how to view detection results and risk reports in the AI Guardrails console.
Detection result query
Result query
You can view the moderation results to analyze the most frequent violation types in the moderated text.
Log on to the AI Guardrails console.
On the page, you can view the moderated text, hit tags, and request time.
You can filter the results by setting search conditions, such as a time range, request ID, text, and tags. You can query data from the last 30 days. The Result Query page stores a maximum of 50,000 records. If you require more storage, save the results returned from API calls.
Find a specific text entry and in the Operation column, click Details to view the detailed moderation information for the text.
Risk report
The risk report shows the Trend Statistics and Risk Distribution of calls for Content security detection, Sensitive content detection, and Prompt injection detection.
On the page, select the Content security detection, Sensitive content detection, or Prompt injection detection tab.
You can view the Trend statistics and Risk Distribution of calls by day, month, account, or service. The Risk Distribution statistics include a list of the top five hit tags and all hit tags.