Skip to content
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

Assistant add Gauge UI to machine generated text #686

Open
wants to merge 8 commits into
base: beta-master
Choose a base branch
from

Conversation

rosannamilner
Copy link
Collaborator

@rosannamilner rosannamilner commented Dec 10, 2024

Adding GaugeChartResult to Machine Generated Text

  • removed left over credibility signal states which aren't used here anymore
  • issue with arcsLengths: I can't seem to get the right DETECTION_THRESHOLDS for the arcs I need to display, so I have added sending predefined arcsLengths as props
    • DETECTION_THRESHOLDS = { THRESHOLD_1: 5.0, THRESHOLD_2: 50.0, THRESHOLD_3: 95.0, }
    • arcsLength = [0.05, 0.45, 0.45, 0.05]
  • needed to edit CustomAlertScore to allow for new detectionType = "machine_generated_text"
  • needed to edit CustomAlertScore to allow for align="start" as text was centered

Questions

  • please let me know if there's any wording I can change/improve
  • currently, when the result is successful the accordion remains closed, I could set it to be open when loaded successfully? (Same for the Previous Fact Checks)

Example link: https://www.breitbart.com/news/suspect-arrested-after-hunt-for-us-insurance-chiefs-killer/
image (87)
image (88)

@rosannamilner rosannamilner changed the title Assistant/mgt add gauge UI Assistant add Gauge UI to machine generated text Dec 10, 2024
@rosannamilner rosannamilner marked this pull request as ready for review December 10, 2024 15:44
Copy link
Collaborator

@jmafoster1 jmafoster1 left a comment

Choose a reason for hiding this comment

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

This looks very nice! 😍

@Sallaa
Copy link
Contributor

Sallaa commented Dec 11, 2024

Hi @rosannamilner, this looks good!

Did you get any feedback from KInIT on this, especially on the probability scores, the scale and their meanings?

Partners in Vera often have different wordings and / or scales for the results of their models and I just want to check that they have been consulted before merging this

@rosannamilner
Copy link
Collaborator Author

Hi @rosannamilner, this looks good!

Did you get any feedback from KInIT on this, especially on the probability scores, the scale and their meanings?

Partners in Vera often have different wordings and / or scales for the results of their models and I just want to check that they have been consulted before merging this

No I haven't, good point - I will do this now!

@rosannamilner
Copy link
Collaborator Author

rosannamilner commented Dec 11, 2024

Kalina has suggested removing the bottom "Hide detection details" entirely as it's duplicated information in this case. Also, there is extra whitepace around the gauge as compared to in Deepfake Image/Video as the image/video would sit on the left hand side of the GaugeChartResult component.

@Sallaa Please let me know your thoughts on this and I can change the UI now, or we can leave this for a future improvement down the line if not too important now?

@Sallaa
Copy link
Contributor

Sallaa commented Dec 13, 2024

Yes, it makes sense to remove the details if it does not add any information to the user. I'm also fine if the UI is not pixel perfect. We can ship it as it is (once you've collected KInIT feedback and scale) and adjust it later as we also collect feedback from users

@rosannamilner
Copy link
Collaborator Author

Yes, it makes sense to remove the details if it does not add any information to the user. I'm also fine if the UI is not pixel perfect. We can ship it as it is (once you've collected KInIT feedback and scale) and adjust it later as we also collect feedback from users

Thanks, removed duplicate information and yes, happy to see what the users think of course. Ivan and KInIT are happy with the text. This is now ready to go.
image (87)

@rosannamilner
Copy link
Collaborator Author

To go with InVID Translations PR: AFP-Medialab/InVID-Translations#109

@Sallaa
Copy link
Contributor

Sallaa commented Jan 7, 2025

@rosannamilner LGTM -- I will merge this once you have resolved the conflicts for the translation PR

@rosannamilner
Copy link
Collaborator Author

@rosannamilner LGTM -- I will merge this once you have resolved the conflicts for the translation PR

Thanks - resolved

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