Autograd for ndarray

Autograd Eigen backend

Tensor computation and autograd with WebGPU acceleration

Forward + reverse-mode automatic differentiation for MathTS rank-N Tensor

PyTorch-like JavaScript tensor/autograd library with CPU training support. Provides Tensor, autograd, nn.Module, Linear, Conv2d, Adam optimizer, and MNIST dataset utilities.

Tiny TypeScript-native tensor library with autograd, compiling to WebGPU. Train small models in the browser without hand-writing kernels, or run a pretrained one frozen for transfer learning.

BuilderForce Agent Memory — engine layer. TypeScript/WGSL Mamba SSM kernels, model blocks, autograd, training, and BPE tokenizer (zero runtime deps, WebGPU-native).

A tiny Autograd engine based off of Andrej Karpathy's micrograd in Python

A simple scalar autograd library for TypeScript

A simple, handrolled autograd implementation package with an MLP wrapper

Minimal neural network from scratch — neuron, layer, network, backpropagation, classical ML, unsupervised, generative models, autograd. No dependencies.

Autograd gpu.js backend

A TypeScript ML framework with Rust native backends (CPU, CUDA, WebGPU) — autograd, tensors, and neural network training at GPU speed.

A tiny reverse-mode automatic differentiation engine in TypeScript

Tensor computation and autograd with WebGPU acceleration

PyTorch-like deep learning library for JavaScript — tensors, autograd, and neural networks in JS

A JavaScript/TypeScript autograd engine with operator overloading, inspired by micrograd

neurova AI core — Tensor (Float32Array) + reverse-mode autograd + nn layers/optim/losses. Pure TS, zero deps.

A small, readable tensor + autograd library that runs inside Pyodide. PyTorch-flavored API, NumPy-backed, designed to be educational source code. Not PyTorch — `import browsergrad_grad as grad`.

Scalar-based reverse-mode automatic differentiation in TypeScript.