Skip to content

FEAT Addition of LookBackScorer which scores using the entire conversation as context. #906

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 30 commits into from
May 9, 2025

Conversation

whackswell
Copy link
Contributor

Description

The LookBackScorer is a new scorer that will evaluate the entire conversation based on a provided set of criteria. The first provided set of criteria is behavior_change.yaml which evaluates behavior of the target due to social engineering.

look_back_scorer.ipynb includes examples of how to use the LookBackScorer.

@bashirpartovi @rlundeen2 @romanlutz

Whitney Maxwell and others added 19 commits April 7, 2025 17:13
…res the entire conversation based on percieved behavior changes.
… and scores the entire conversation based on percieved behavior changes."

This reverts commit 965fbf0.
…ude examples of persuasion and deception files.
Copy link
Contributor

@romanlutz romanlutz left a comment

Choose a reason for hiding this comment

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

Great work! We will need a few unit tests.

Copy link
Contributor

@romanlutz romanlutz left a comment

Choose a reason for hiding this comment

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

Thanks for adding the longer conversation text. This looks great 🙂

@romanlutz romanlutz changed the title FEAT Addition of LookBackScorer which scores using the entire conversation as context. Example code included. FEAT Addition of LookBackScorer which scores using the entire conversation as context. May 9, 2025
@romanlutz romanlutz merged commit 77c095b into Azure:main May 9, 2025
19 checks passed
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.

3 participants