Fix Enzyme forward rule for Krylov methods (#766) #767
Conversation
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
- Implement proper forward differentiation for AbstractKrylovSubspaceMethod algorithms - Previously, forward rule threw 'Algorithm not supported' error for Krylov methods - Reverse rule already had correct implementation for these methods - Add test to verify forward rule no longer throws the specific error Addresses issue SciML#766 where KrylovJL_GMRES failed with Enzyme differentiation. The forward rule now computes dy/dt = A^(-1)(db/dt - dA/dt * y) for Krylov methods. 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <[email protected]>
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Outdated suggestions cannot be applied.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.
Summary
Fixes inconsistency in Enzyme extension where forward rule threw an error for Krylov methods while reverse rule properly supported them.
Problem
Issue #766 reported Enzyme differentiation failing with KrylovJL_GMRES. Investigation revealed:
Solution
Implemented proper forward differentiation for Krylov methods in the forward rule:
y = A⁻¹b, computesdy/dt = A⁻¹(db/dt - dA/dt * y)db - dA * resChanges
Testing
Notes
This fixes the specific inconsistency in LinearSolve.jl's Enzyme extension. The broader issue in #766 involving constant variable handling may require additional work, but this addresses the immediate compilation error with Krylov methods.
🤖 Generated with Claude Code