PinnedInGenerative AIbyHank S·Aug 31My LLM Benchmark Came Out Flawless. That Was My First Warning Sign.How a 0% attack rate led to a much bigger discovery about LLM activation steering and what actually matters in AI research.A response icon1A response icon1
InLevel Up CodingbyHank S·Sep 9We Focus on the 99% in AI Safety. We Really Need to Worry About the 1%.How extreme tail events fan out and how they can potentially be mitigated at pre-training.
InGenerative AIbyHank S·Sep 5Your Agent Setup will Probably Fail in Production. Here is How to Avoid it.How Pydantic helps with functions in tool calling in both ReAct and graph-based paradigms.A response icon2A response icon2
Hank S·Mar 25AI Didn’t Raise the Bar. It Removed the Ladder.AI has made entry-level engineering jobs disappear. That’s not the whole picture — but it’s enough to change the tech scene forever.
InGoPenAIbyHank S·Mar 4From Tokenization to Softmax: Master Transformer BasicsThese 4 concepts are the 20% that drive 80% of your Deep Learning success.
InGoPenAIbyHank S·Jan 31Attention Isn’t All You Need: The Essential Foundations of the Transformer.Autograd is the engine, but the architecture is the plumbing. Deep networks don’t fail because the math is wrong; they fail because the…
InGoPenAIbyHank S·Jan 225 Steps to Memorizing the Autograd Implementation for GoodIf you want to break into a top frontier research lab, knowing how the autograd library functions from the ground up is non-negotiable.
Hank S·Sep 23, 20255 AI Prompts That Will Seriously Save You 20+ Hours a Week (Free Guide)A practical guide to prompt engineering. Use 5 AI prompts to boost productivity, simplify tasks, and unlock your creativity.
