fp16.in is a personal log of AI research and engineering — what a paper claims, what a model actually does when you run it, and what's worth keeping once the hype is stripped out.
{{ typedTagline }}▌
I build LLM and RAG systems for a living — retrieval pipelines, fine-tuned open models, agents that mostly behave. This is where I write down what I learn along the way, minus the parts that don't hold up.
No schedule, no series. Just a note when something's worth reading.