In this paper, we introduce a verified framework for defining and composing sparse tensor formats. We extend the ATL tensor language and scheduling framework, which formerly could only express dense tensor kernels. We define a levelized abstraction to describe per-dimension tensor formats via their encoding routines, access and iteration functions, and formal properties enforcing soundness of the sparse structures as representations of the original dense tensors. Using this abstraction, we compositionally define format-agnostic, multidimensional compression and decompression functions that are used to express the top-level soundness theorem for these abstract sparse tensor formats. We then use this soundness theorem as an adjoint-pair rewrite theorem to introduce sparse data structures and iteration into a dense tensor kernel via the existing scheduling-rewrite framework of ATL. Overall, we are able to start with a program computing over dense operands and derive a proven semantically equivalent, optimized program computing over sparse structures. We further prove a minimal set of instances of the level-format abstraction, which can be composed and passed as parameters to compression to capture a broad range of canonical, multidimensional tensor-compression formats.

@article{paperbot3695,
  title = {A Mechanized Algebra of Verified Data Structures for Optimizing Sparse Tensor Programs},
  author = {Amanda Liu and Gilbert Louis Bernstein and Shoaib Kamil and Adam Chlipala and Jonathan Ragan-Kelley},
  journal = {Proceedings of the ACM on Programming Languages},
  volume = {10},
  number = {PLDI},
  year = {2026},
  doi = {10.1145/3808261}
}