Deep Dive

In-depth technical articles on AI, GPU inference, and developer tools

LLM

My RAG's Model Had Already Read the Books - Per-Book Verdicts, a Fake Regression, and Catching a 9B Leaking Prior Knowledge

I run a local RAG over all seven Harry Potter novels with a 9B model that knows the franchise cold. Re-running my eval o...

 LLM

Next-Gen LLMs: Deep Dive into Compact, High-Speed Models and Temporal Reasoning – Gemini 3.1 Flash-Lite, GPT-5.4 mini/nano

Today's Highlights Hello, fellow personal developers and AI researchers! In today's tech...