--- title: "#mlx articles | Hashnode" description: "Explore 22 articles tagged #mlx on Hashnode. Tutorials, guides, and how-tos from the community." url: "https://hashnode.com/tag/mlx" --- Tag # #mlx 22 posts · 0 followers [Martin Bowling](https://hashnode.com/@martinbowling) · [Oct 27](https://hashnode.com/posts/the-celebrity-neuron-is-not-a-neuron/6ae0ecbf22d9441869e7cab6) ### [The Celebrity Neuron Is Not a Neuron](https://martinbowling.com/the-celebrity-neuron-is-not-a-neuron?utm_source=hashnode&utm_medium=feed) I tried to make Qwen3.5-4B identify as Einstein, Curie, Newton, Taylor Swift, and Keanu Reeves. The scientists worked. The entertainers didn’t. 5 min read · [martinbowling.com](https://martinbowling.com/the-celebrity-neuron-is-not-a-neuron?utm_source=hashnode&utm_medium=feed) [Martin Bowling](https://hashnode.com/@martinbowling) · [Oct 26](https://hashnode.com/posts/i-audited-the-viral-qwen-neuron-zoo/6adf95f334d8bf647f016915) ### [I Audited the Viral Qwen Neuron Zoo](https://martinbowling.com/i-audited-the-viral-qwen-neuron-zoo?utm_source=hashnode&utm_medium=feed) I tested the viral one-neuron personas on Qwen3.5-4B. 16 of 22 survived a one-digit fix, one was impossible, and three neurons become the Golden Gate Bridge. 10 min read · [martinbowling.com](https://martinbowling.com/i-audited-the-viral-qwen-neuron-zoo?utm_source=hashnode&utm_medium=feed) [Martin Bowling](https://hashnode.com/@martinbowling) · [Oct 26](https://hashnode.com/posts/three-neuron-bangers/6adfe6ccbaf492ef8de66762) ### [Three-Neuron Bangers](https://martinbowling.com/three-neuron-bangers?utm_source=hashnode&utm_medium=feed) Five combos that fuse into something none of the neurons does alone — plus two cursed loops and one random control that does nothing. 5 min read · [martinbowling.com](https://martinbowling.com/three-neuron-bangers?utm_source=hashnode&utm_medium=feed) [Haithem Slimi](https://hashnode.com/@haithem-slimi) · [Oct 25](https://hashnode.com/posts/teaching-a-small-local-model-to-think-in-differentials/6ade340c34e352aaff851fec) ### [Teaching a Small Local Model to Think in Differentials](https://nightthoughts.me/teaching-a-small-local-model-to-think-in-differentials?utm_source=hashnode&utm_medium=feed) A DDXPlus and QLoRA experiment on Apple Silicon, from held-out evaluation to oMLX deployment 31 min read · [nightthoughts.me](https://nightthoughts.me/teaching-a-small-local-model-to-think-in-differentials?utm_source=hashnode&utm_medium=feed) [Mihai\_LeanZero](https://hashnode.com/@mihaileanzero) · [Oct 16](https://hashnode.com/posts/best-local-image-generation-model-on-mac-ernie-vs-flux/6ad28472639d2410262e5df9) ### [Best Local Image Generation Model on Mac: ERNIE vs FLUX](https://leanzero.hashnode.dev/best-local-image-generation-model-on-mac-ernie-vs-flux?utm_source=hashnode&utm_medium=feed) The best local image generation model on a Mac, measured: six open models blind-scored and timed on an M3 Ultra, plus which licenses allow commercial use. 29 min read · [leanzero.hashnode.dev](https://leanzero.hashnode.dev/best-local-image-generation-model-on-mac-ernie-vs-flux?utm_source=hashnode&utm_medium=feed) [Haithem Slimi](https://hashnode.com/@haithem-slimi) · [Oct 9](https://hashnode.com/posts/complete-guide-always-on-mac-server-with-mlx-omlx-qwenpaw/6ac956148e77433892c6107f) ### [Complete Guide: Always-On Mac Server with MLX, oMLX & QwenPaw](https://nightthoughts.me/complete-guide-always-on-mac-server-with-mlx-omlx-qwenpaw?utm_source=hashnode&utm_medium=feed) Headless MLX inference with oMLX, QwenPaw, Caddy, Tailscale, and Cloudflare — verified end to end. 50 min read · [nightthoughts.me](https://nightthoughts.me/complete-guide-always-on-mac-server-with-mlx-omlx-qwenpaw?utm_source=hashnode&utm_medium=feed) [Dragos Roua](https://hashnode.com/@dragosroua) · [Oct 3](https://hashnode.com/posts/video-how-to-set-up-local-ai-on-a-16gb-macbook/6ac04a9fa62469fb09002d2d) ### [\[ Video \] How To Set Up Local AI on a 16GB MacBook](https://dragosroua.hashnode.dev/video-how-to-set-up-local-ai-on-a-16gb-macbook?utm_source=hashnode&utm_medium=feed) Setting up local AI on a 16GB MacBook involves using optimized binaries for the Metal architecture. Start with MLX-serve. Optionally OpenWebUI 4 min read · [dragosroua.hashnode.dev](https://dragosroua.hashnode.dev/video-how-to-set-up-local-ai-on-a-16gb-macbook?utm_source=hashnode&utm_medium=feed) [Mihai\_LeanZero](https://hashnode.com/@mihaileanzero) · [Sep 23](https://hashnode.com/posts/goose-30-local-models-mlx-and-leanzero-link/6ab3715b6a4187d67aac2dae) ### [Goose 3.0: Local Models, MLX and LeanZero Link](https://leanzero.hashnode.dev/goose-30-local-models-mlx-and-leanzero-link?utm_source=hashnode&utm_medium=feed) Goose Swarm 3.0 runs local models on its own MLX engine, joins your Macs with LeanZero Link and serves an OpenAI-compatible goose endpoint. 26 min read · [leanzero.hashnode.dev](https://leanzero.hashnode.dev/goose-30-local-models-mlx-and-leanzero-link?utm_source=hashnode&utm_medium=feed) [Mihai\_LeanZero](https://hashnode.com/@mihaileanzero) · [Sep 19](https://hashnode.com/posts/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably/6aaec3c26eadb167d1790c1e) ### [Qwen3-Coder-Next + MXFP8: The 128GB Local LLM That Runs Predictably](https://leanzero.hashnode.dev/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably?utm_source=hashnode&utm_medium=feed) The MXFP8 MLX quant of Qwen3-Coder-Next (80B total, 3B active, 262K context) on a 128GB Mac: LM Studio 0.4.1 setup for Claude Code. 34 min read · [leanzero.hashnode.dev](https://leanzero.hashnode.dev/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably?utm_source=hashnode&utm_medium=feed) [2](https://hashnode.com/posts/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably/6aaec3c26eadb167d1790c1e) [Mihai\_LeanZero](https://hashnode.com/@mihaileanzero) · [Sep 19](https://hashnode.com/posts/mlx-vs-gguf-on-apple-silicon-benchmarking-the-same-local-model-two-ways/6aaeb06506ef87fd984a3ef5) ### [MLX vs GGUF on Apple Silicon: Benchmarking the Same Local Model Two Ways](https://leanzero.hashnode.dev/mlx-vs-gguf-on-apple-silicon-benchmarking-the-same-local-model-two-ways?utm_source=hashnode&utm_medium=feed) One 27B model as GGUF and MLX, 30 app builds graded by running them. Each worked 13 of 15, speed was a wash, and MLX was steadier: the safe default. 13 min read · [leanzero.hashnode.dev](https://leanzero.hashnode.dev/mlx-vs-gguf-on-apple-silicon-benchmarking-the-same-local-model-two-ways?utm_source=hashnode&utm_medium=feed)