AVB (@neural_avb)

JEV actually doesn't even run deterministically Exact same prompts give you different probabilities when you run it multiple times The ORDER of the choices DRASTICALLY changes the output probs I am more and more confused by the their "no-hallucination" claim <span class="quote-card" data-quote-i...

My ~40 minute tutorial about building a feature-rich CODING AGENT HARNESS from scratch, no frameworks Tools, bash, file reads, writes, patch updates, cache-aware prompting, compaction, context mgmt, late injections, TODOs, plans, permissions, sandboxes, subagents... 👇🏼 ...

The promise of <b>knowledge distillation</b> is quite interesting. It is a machine learning technique that lets you distill, or transfer, knowledge from a more capable <b>teacher network</b> into a less capable<b> student network</b>....

I made a 1.5 hour tutorial on Context Engineering with DSPy. Goes from the basics to the more crazy stuff. Repo and related resources in the comments. ...

RLMs explained and implemented from first principles in a (50 minute) nutshell <span class="quote-card" data-quote-index="0"></span>...

So there are basically two ways people make a LLM "smarter": 1. give it <b>more parameters</b> (a bigger brain), 2. or, give it <b>more compute/data</b> to train on (a longer education). Both work, but both are expensive!...