Scope keyed-reduction scratch leaves by stage - #26
Merged
Merged
Conversation
PraneethMerugu
marked this pull request as ready for review
September 24, 2026 04:40
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A sequence of keyed reductions with different tuple key layouts reused the first stage scratch leaf names, causing the second stage to receive the wrong component arrays and fail on signed owner keys. Scope component leaf names to their physical stage. The same observable mixed-key sequence regression runs on CPU and real Metal. This unblocks CorePotts PR37 fixed-domain-owner relation reconstruction without changing the keyed-reduction execution law. Local full owner CPU: 1947/1947. Local full ordinary real-Metal suite passed with scalar indexing disabled, including the new mixed-key witness. Hosted required checks remain merge authority.