fix(sleep): align dream contrast with gate metric - #211
Open
Bogdan (Dan) Baciu (bogdanbaciu21) wants to merge 1 commit into
Open
fix(sleep): align dream contrast with gate metric#211Bogdan (Dan) Baciu (bogdanbaciu21) wants to merge 1 commit into
Bogdan (Dan) Baciu (bogdanbaciu21) wants to merge 1 commit into
Conversation
Contributor
Author
|
Got the PR up, will fill in details ASAP on the contribution and technical details + test suite coverage. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes a mismatch in SkillOpt’s dreaming path: contrastive reflection always ranked rollouts by hard score, even when validation used soft or mixed scoring. Tied hard outcomes with different partial-credit quality therefore produced no contrast and no dream learning.
This change uses the configured gate metric for contrast selection, preserves hard scoring as the default, and adds regression tests for hard, soft, mixed, and configuration propagation.
Validation: 108 focused tests passed. Full suite: 932 passed, 8 skipped. Ruff passed.