Fast matrix multiplication in your browser

WASM matmul solver kernel for @btx-tools/challenges-sdk — byte-exact port of btxd's matmul service-challenge PoW

GPU-first numerical computing framework — Runtime, 32 Compute primitives (Map, Reduce, Scan, Sort, FFT, MatMul, etc)

WebGPU matmul solver kernel for @btx-tools/challenges-sdk — byte-exact WGSL port of btxd's matmul service-challenge PoW

WGSL kernel catalog for browser-side machine learning: matmul, softmax, layernorm, attention, and friends. Each kernel ships with a JS reference for conformance + fallback. Independent of any tensor framework.

Rehype plugin for rendering ML tensor/memory diagrams (shape/strides, matmul tiling, broadcasting) as inline SVG

A fast and efficient WebGPU powered implementation of matrix multiplication

Pool-agnostic browser mining client for BTX matmul service-challenge shares (WebGPU → WASM → pure-JS)

DXO deep learning - TypeScript entry (developer preview, API unstable)

Principal Components Analysis in javascript

Shared WebGPU foundation for MathTS — device/context lifecycle, buffer pool, and a generic shader manager (no domain kernels)

TypeScript SDK for BTX service challenges — chain-anchored proof-of-work admission control for APIs, agent gateways, and form submissions

Tool to convert an ONNX graph into a data-flow graph (DFG), decomposing its high-level operations into low-level operations and performing a set of optimizations.

DXO WebGPU/WASM runtime - createRuntime + CPU fallback (developer preview, API unstable)

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

Tensor for JavaScript

On-device language identification for short text, across 84 languages. Pure JavaScript: no wasm, no inference runtime.

Numerical methods for web

No description provided.

NumType — n-dimensional arrays for TypeScript with compile-time shape checking (research spike)