Skip to content

Feat: Report specificity and confidence intervals in analytics - #3586

Closed
🐦[T3T]Berserker 🐦 🐍 (TrinityBerserker) wants to merge 1 commit into
microsoft:mainfrom
TrinityBerserker:feat/analytics-specificity-confidence-intervals
Closed

🐦[T3T]Berserker 🐦 🐍 (TrinityBerserker) wants to merge 1 commit into
microsoft:mainfrom
TrinityBerserker:feat/analytics-specificity-confidence-intervals

Conversation

@TrinityBerserker

This PR resolves #3579 by reporting scorer specificity alongside attack success rates and adding confidence intervals to outcome statistics.

Changes:

Added specificity to ObjectiveScorerMetrics and computed it in _compute_metrics.

Added confidence_interval_lower and confidence_interval_upper to AttackAnalyticsStatistics.

Implemented the Wilson score interval to calculate confidence intervals for the success rate in compute_outcome_statistics.

All existing unit tests pass (114 tests in test_analytics.py and 24 in test_derivation.py).

Fixes #3579

@TrinityBerserker

Copy link
Copy Markdown
Author

@microsoft-github-policy-service agree

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

FEAT: Report confidence intervals and scorer specificity alongside attack success rates in analytics

2 participants