neuron-explorer
Steering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout, browse the results
Batched QR decomposition, symmetric eigendecomposition (eigh) and SVD on Apple Silicon GPUs: Metal kernels for MLX and PyTorch, with C++, Swift and C APIs.
Developer toolsC · C++GitHub @c0rmacOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
QR, symmetric eigendecomposition and SVD for batches of matrices on Apple Silicon GPUs, for MLX and PyTorch. Metal kernels per regime, routed to the GPU or the CPU by policies measured per Mac. C++, Python (pip: metal-linalg, metal-linalg-torch), Swift and C.
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
QR decomposition, symmetric eigendecomposition (eigh) and singular value
decomposition (SVD) for batches of matrices on Apple Silicon GPUs, for
MLX, PyTorch and
plain float buffers. A C++ library, installed with Homebrew or built from
source inside your own project, with Python packages for mlx.core arrays
and for torch tensors, a C API, and a Swift package (on [Float], and on
mlx-swift's MLXArray).Read the full README ↗The full README is unavailable right now. Read it at the original link. The saved excerpt is shown below.
QR decomposition, symmetric eigendecomposition (eigh) and singular value
decomposition (SVD) for batches of matrices on Apple Silicon GPUs, for
MLX, PyTorch and
plain float buffers. A C++ library, installed with Homebrew or built from
source inside your own project, with Python packages for mlx.core arrays
and for torch tensors, a C API, and a Swift package (on [Float], and on
mlx-swift's MLXArray).Read the full README ↗