PinnedHamman Samuel, PhD·21h agoGenAIOpsHow do you ship a new prompt, switch models while the app is live, and make sure an agent still does what the user asked?
PinnedInGenerative AIbyHamman Samuel, PhD·Sep 30From Navier–Stokes to SciCode: Testing AI at the Scientific FrontierCan an AI system understand a research-level scientific problem and turn that into working numerical code?
PinnedInGoPenAIbyHamman Samuel, PhD·Sep 4LLM inference engines explainedLLM inference engines explained: llama.cpp, HFT, vLLM, SGLang, CPU, GPU, and vendor-specific stacks.
PinnedInGoPenAIbyHamman Samuel, PhD·Oct 1, 2025Guide to Fine-Tuning LLMsWe’ll walk through fine-tuning an open-source LLM and adapting it to our specific needs.
InGoPenAIbyHamman Samuel, PhD·Nov 6Getting to Know Meta’s Muse Wonder LLMMuse Wonder caught my attention at the top of an interesting software engineering benchmark.
InGoPenAIbyHamman Samuel, PhD·Oct 29AI Probes: Keeping your Chatbot on TrackFour conversational probes let you keep your chatbot on track: canary, honeypot, checksum, and echo
InGoPenAIbyHamman Samuel, PhD·Oct 13Maturing LLM PromptsFrom text strings to structured prompts using DSPy
InGoPenAIbyHamman Samuel, PhD·Oct 1System One Models in Practice — Jev, Laya, Drex, Julia-1 and moreThis article covers System One models like Jev, Laya, Drex, Julia-1, and also Nimble, Tev1, and Kev as LLM-derived decision models.A response icon3A response icon3