Generate deterministic colors across CPU and GPU with SplitMix64 PRNG.
npx skills add https://github.com/plurigrid/asi --skill mlx-jax-splitmix
Please help me install this Agent Skill. Skill: mlx-jax-splitmix Source: https://github.com/plurigrid/asi/tree/main/skills/mlx-jax-splitmix Command: npx skills add https://github.com/plurigrid/asi --skill mlx-jax-splitmix
Provides deterministic color generation across CPU/GPU using SplitMix64, with JAX-like API and MLX acceleration on Apple Silicon.
Run the MLX/JAX demo to derive colors for a seed, e.g., 1069, across 100 steps.
High-intent search queries and answers about installing and using this skill.
Deterministic color generation uses a PRNG—SplitMix64—to derive identical color outputs from identical seeds across devices and runs. This Skill applies SplitMix64 with MLX acceleration on Apple Silicon, ensuring consistent hues whether you compute on CPU or GPU, enabling reproducible workflows across machines.
Yes. This Skill provides a JAX-style SplitMix64 PRNG implemented with MLX on Apple Silicon, enabling GPU-accelerated deterministic color derivation. You get functional PRNG semantics—independent streams and key splitting—with native MLX performance on Apple hardware.
Use deterministic seed derivation with SplitMix64 to generate color streams that remain identical across CPUs, GPUs, and machines. This Skill implements batch processing and GPU acceleration with MLX, allowing you to derive consistent hues for multi-device workflows without synchronization overhead.
No. This Skill provides a JAX-style API but runs on MLX—Anthropic's framework for Apple Silicon. You get the familiar JAX semantics for functional PRNG and deterministic color derivation without requiring JAX, optimized directly for Apple GPU hardware.
SplitMix64 uses key splitting to generate independent PRNG streams from one seed. Each stream produces a different sequence of deterministic values, allowing you to derive separate color sequences from a single seed while maintaining reproducibility across runs and devices.
Deterministic color generation is platform-specific: this implementation targets Apple Silicon with MLX. Scale is bounded by batch processing capacity and available GPU memory. Seed derivation remains consistent across runs, but color mappings depend on your chosen palette and bit-depth constraints.
hhhi21gQuantize llama.cpp models to 2-8 bit depths using C/C++.
devMoezGenerate browser-ready p5.js sketches and export them as PNG, GIF, MP4, or SVG.
plurigridCoordinate distributed LLM inference across Apple Silicon devices with auto peer discovery.
plurigridGenerate deterministic #RRGGBB colors from metadata hashes for IsUMAP visualizations.
