Repository navigation
Home: show laya:typed-decisions and jeb in the typed-decisions scoreboard - #85
Merged
Merged
Conversation
…oard Both were left out because they trained on the dataset's train split, which made TypeSafe's reported Jev look ahead of every model. They now appear with an asterisk and a footnote (the bench's convention), the axis runs to 1.00 so jeb's 0.802 fits, the name column is wide enough for laya:typed-decisions, and the tick labels hide below 360 px where they overlapped. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The home scoreboard left out the two models trained on typed-decisions' train split, so TypeSafe's reported 0.738 for Jev sat above every row. They now appear, marked:
jeb:27b *(0.802) andlaya:typed-decisions *(0.767), with the footnote "Trained on its train split, so these models have seen questions like these" and the same text on hover, as the bench marks contamination.Checked at 1280, 640, 390 and 320 px;
npm run buildpasses.🤖 Generated with Claude Code