Skip to content

DSB #145


DSB has arrived. Don’t want to waste your time, so go read about Mamba and state space models, or watch another tutorial video by Andrej Karpathy!

And as always, enjoy your reading.

Science –  Is the Mamba framework, based on state space model (SSM) architecture, better and quicker in language processing compared to Transformers? This overview might provide the answer. You could also try this article about Mamba basics and SSM. –  This goes beyond comprehensive – it’s a ridiculously exhaustive list of resources for (aspiring) full-stack programmers. But you should give it a shot. – Obviously, I like Andrej Karpathy; his videos are amazing. If you want to get familiar with LLM, he is the one. This video provides an in-depth look at how to construct a tokenizer. – Forget prompts and manual prompt engineering; DSpy gives you dynamic, programmatic control over your LLM pipeline. Also watch this introductory video and learn more.

Pop – Not even curated repositories are 100% safe. JFrog identified about 100 malicious AI models on Hugging Face, illustrating that advanced security measures can still be circumvented. – I like this article. Being a data-driven company does not mean that you can get rid of human intuition and experience. – What does being a data scientist in a corporation mean? It means being proactive, understanding the product, and contributing to both strategy and development. – Usually, I don’t like lists like this one, but it is interesting and insightful to see what the listed fintechs are working on.


Be First to Comment

Leave a Reply