Skip to content

[AI] add static indexing for discrete/constants/nonnumeric#4632

Open
SebastianM-C wants to merge 1 commit into
SciML:masterfrom
SebastianM-C:fix/enzyme-nonnumeric-template-inference
Open

[AI] add static indexing for discrete/constants/nonnumeric#4632
SebastianM-C wants to merge 1 commit into
SciML:masterfrom
SebastianM-C:fix/enzyme-nonnumeric-template-inference

Conversation

@SebastianM-C

@SebastianM-C SebastianM-C commented Jun 16, 2026

Copy link
Copy Markdown
Member

This adds StaticBufferIndex to perform indexing in the type domain for the discrete/constants/nonnumeric portions.
This mainly helps Enzyme's type analysis for use cases that include nonnumeric parameters, like neural networks.

I'm not sure if we would have any cases where static indexing of those portions would be unreasonable. Besides that, I mainly care about the nonnumeric case, the others were added for consistency, but I can drop that if it would create issues.

Checklist

  • Appropriate tests were added
  • Any code changes were done in a way that does not break public API
  • All documentation related to code changes were updated
  • The new code follows the
    contributor guidelines, in particular the SciML Style Guide and
    COLPRAC.
  • Any new documentation only uses public API

Additional context

See SciML/SciMLSensitivity.jl#1477 for more details.

This adds StaticBufferIndex to perform indexing
in the type domain for the
discrete/constants/nonnumeric portions.

This mainly helps Enzyme's type analysis
for use cases that include nonnumeric
parameters, like neural networks.
See SciML/SciMLSensitivity.jl#1477 for more details.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@SebastianM-C SebastianM-C changed the title add static indexing for discrete/constants/nonnumeric [AI] add static indexing for discrete/constants/nonnumeric Jun 16, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant