Add optional weight canonicalization before Phase 2 similarity signals#9
Open
karthikarunapuram8-dot wants to merge 1 commit intocisco-ai-defense:mainfrom
Conversation
karthikarunapuram8-dot
force-pushed
the
feat/weight-canonicalization
branch
from
6c548a6 to
3351b3a
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds an optional canonicalization pass for weight-space provenance signals to reduce false negatives from transformer permutation and scale symmetries. The pass aligns attention heads and MLP channels where architecture metadata permits, normalizes scale-sensitive channels, preserves default behavior, and reports canonicalization metadata in JSON output.