Safety Monitor
Collinear empowers your business with a dashboard that tracks key metrics to ensure your AI performs safely, accurately, and reliably.
π₯ Interactive Walkthrough
Explore the Safety Dashboard in action with an interactive demo:
π Key Metrics Overview
Track the most important indicators of model behavior in real time:
-
Total Volume Total number of queries processed by real-time judges.
-
Prompt Safety Violation Rate Percentage of unsafe or harmful prompts submitted by users.
-
Response Safety Violation Rate Percentage of model responses that violate safety guidelines.
-
False Refusal Rate Percentage of safe user prompts that the AI incorrectly refuses to answer.
π Critical Summaries
Gain high-level visibility into the systemβs behavior:
-
Total Queries vs. Flagged Queries A time-series graph comparing overall query volume to those flagged by the system.
-
Risk Levels Breakdown of queries by assessed risk level.
-
Flagged Categories Categorization of flagged queries to identify common safety issues.
π Live Data
Dive into raw safety data with the live table, including:
- ID β Unique identifier for each query.
- Conversation Prefix β The initial part of the user input.
- Model Response β The AIβs reply to the prompt.
- Judge Score β Evaluation score assigned by the judge model.
- Categories β Tags describing the nature of the issue, if any.
- Actions β What actions were taken in response (e.g., flagged, passed, needs review).
Was this page helpful?