Skip to content

Introduce Content Safety evaluators - #6223

Merged
shyamnamboodiripad merged 4 commits into
dotnet:mainfrom
shyamnamboodiripad:rai
Apr 4, 2025
Merged

Introduce Content Safety evaluators#6223
shyamnamboodiripad merged 4 commits into
dotnet:mainfrom
shyamnamboodiripad:rai

Conversation

@shyamnamboodiripad

@shyamnamboodiripadshyamnamboodiripad commented Apr 1, 2025•
edited
Loading

Copy link
Copy Markdown
Contributor

The new evaluators will ship in a new Microsoft.Extensions.AI.Evaluation.Safety package.

Also includes the following public API changes:

  • Add a Metadata dictionary on EvaluationMetric.
  • Make EvaluationMetric.Diagnostics nullable.
  • Convert instance functions on some (fully mutable) result types to extension methods in the same namespace.

Some more work remains before this can be merged:

  • Add a couple more evaluators that are currently missing
  • Add more documentation comments for the new evaluators and metrics
  • Display metadata for individual metrics in the report
  • Some miscellaneous clean up
  • Add few more tests

Fixes#5937

Microsoft Reviewers: Open in CodeFlow

@shyamnamboodiripad
shyamnamboodiripad requested review from a team as code ownersApril 1, 2025 11:03
@github-actionsgithub-actionsBot added the area-ai-eval Microsoft.Extensions.AI.Evaluation and related label Apr 1, 2025
peterwald
peterwald approved these changes Apr 1, 2025
View reviewed changes

@peterwaldpeterwald left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

:shipit:

@shyamnamboodiripad
shyamnamboodiripadforce-pushed the rai branch 2 times, most recently from 058b183 to 1fde648CompareApril 1, 2025 20:54
RussKie
RussKie reviewed Apr 1, 2025
View reviewed changes
Comment threadsrc/Libraries/Microsoft.Extensions.AI.Evaluation.Console/README.md
@shyamnamboodiripad
shyamnamboodiripadforce-pushed the rai branch 5 times, most recently from 40fb1c6 to 00644d3CompareApril 3, 2025 14:23
@shyamnamboodiripad
API changes
686f914
Includes the following API changes:
- Add a Metadata dictionary on EvaluationMetric
- Make EvaluationMetric.Diagnostics nullable
- Convert instance functions on some (fully mutable) result types to extension methods in the same namespace
@shyamnamboodiripad
Change boolean metric UI representation in metric card from Pass / Fa…
a998622
…il to Yes / No
@shyamnamboodiripad
Scaffolding for MEAI.Evaluation.Safety
31c0ebc
@shyamnamboodiripad
Introduce Content Safety evaluators
60b72b1
@shyamnamboodiripad
shyamnamboodiripad enabled auto-merge (squash) April 4, 2025 11:10
@shyamnamboodiripad
shyamnamboodiripad merged commit 68d1661 into dotnet:mainApr 4, 2025
@github-actionsgithub-actionsBot locked and limited conversation to collaborators May 5, 2025
Sign up for freeto subscribe to this conversation on GitHub. Already have an account? Sign in.

Reviewers

@peterwaldpeterwaldpeterwald approved these changes
+1 more reviewer
@RussKieRussKieRussKie approved these changes
Reviewers whose approvals may not affect merge requirements

Assignees

@shyamnamboodiripadshyamnamboodiripad

Labels

area-ai-evalMicrosoft.Extensions.AI.Evaluation and related

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

[AI Evaluation] Introduce RAI evaluators

3 participants

@shyamnamboodiripad@peterwald@RussKie