Faster KNN search in Manticore: 2-pass HNSW, batched distances, and AVX-512(medium.com)52 points by snikolaev 4 months ago | 3 comments
COBOL to Kotlin via Formal Models (IR and Alloy and Golden Master)(marcoeg.medium.com)47 points by marcoeg 11 months ago | 14 comments
Compiling LLMs into a MegaKernel: A path to low-latency inference(zhihaojia.medium.com)291 points by matt_d 1 year ago | 74 comments
I ran Gemma 4 as a local model in Codex CLI(blog.danielvaughan.com)222 points by dvaughan 6 months ago | 94 comments