Release notes:

v0.16.3 — 2026-02-22

A patch release with expanded Mooncake support, improved GPU compatibility, and documentation updates.

Highlights

• Expanded Mooncake AD rules for broader automatic differentiation coverage (#356).
• Adapt support for BraidingTensor, enabling GPU array adaptation workflows (#374).
• Additional upstream and CUDA compatibility fixes, including Base.ones/zeros accepting CuArray (#373).

Notable changes

Added

• Expanded set of Mooncake AD rules for more comprehensive differentiation support (#356).
• Adapt.jl support for BraidingTensor, allowing it to be adapted to GPU arrays (#374).

Changed

• Documentation improvements and updates across multiple sections (#345).

Fixed

• Base.ones and Base.zeros now accept CuArray storage types for CUDA tensors (#373).
• Additional small fixes for upstream Julia and CUDA compatibility (#373).

Upgrade notes

• This is a patch release; the public API remains compatible with v0.16.2.
• Mooncake rule expansion may improve differentiation of previously unsupported operations.

Full changelog