zigllm is an educational, book-shaped Zig codebase that walks you through transformer architectures across six progressive layers: tensors, linear algebra, neural primitives, transformers, models, and inference.https://zigllm.cognisoc.com/https://zigllm.cognisoc.com/blog/sampling-loop-becomes-generation/https://zigllm.cognisoc.com/blog/sampling-loop-becomes-generation/Generation is not a single decision. It's a loop, a probability distribution, and a choice about how greedy or how wild to be. zigllm's Layer 6 makes the loop visible — and shows you why temperature, top-k, and top-p exist.Thu, 02 Apr 2026 00:00:00 GMThttps://zigllm.cognisoc.com/blog/tensors-lingua-franca/https://zigllm.cognisoc.com/blog/tensors-lingua-franca/Tensors aren't a 'data structure.' They're a language. Every layer of zigllm — from raw memory to a sampled token — speaks it, and the shape of each tensor tells you what stage of the pipeline you're in.Wed, 18 Mar 2026 00:00:00 GMThttps://zigllm.cognisoc.com/blog/matmul-by-hand/https://zigllm.cognisoc.com/blog/matmul-by-hand/Matrix multiplication is the heart of a transformer. Writing it by hand in Zig — strides, SIMD, K-quantization — is where 'I know how this works' stops being a claim and starts being legible.Wed, 04 Mar 2026 00:00:00 GMT