Repository navigation
Record cost and turns per arm in the eval baseline - #31
Merged
Merged
Conversation
The score alone hides a skill that reaches the same result at a higher cost. twd-mock-component on Opus scores 1.00 in both arms and costs 57% more with the plugin.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds a second baseline table with the mean cost and turns per run, with and without the plugin, for Opus, Sonnet and Haiku, from the runs behind the existing score table.
The score alone hides a skill that reaches the same result at a higher cost:
twd-mock-componenton Opus scores 1.00 in both arms and costs 57% more with the plugin. The note says where to readcostUsdandturnsper run inaggregate-result.json.