atreeleaf

haha, fwiw this is (surprisingly) still very effective. I imagine we will also start seeing ads on robotaxis soon once they figure out they have "unexplolited" eyeballs

Often times, we lose sight in the magic behind neural nets and get discouraged while learning about some new modeling architecture. I felt this personally when studying the transformer architecture and trying to grasp its complex components such as the multi-headed attention and its positional encoding scheme. I wrote a post here describing the magic behind these magnificent networks, to remind us not to get intimidated! At the end of the day, we are still just simply feeding data into the model and allowing back propagation to work its magic.

Here it is. Cheers!

https://gradiently.io/the-magic-behind-neural-networks/

Hello, friends and strangers on the internet. Happy Thanksgiving!!

I'm grateful for the wealth of information out on the internet that's enabled people around the world to learn anything their curiosity leads them to.

In 1948, Claude Shannon, the father of information theory formalized a mathematical framework in quantifying the amount of information needed to accurately send/receive messages, determined by the degree of "uncertainty" a message could contain.

In his paper, he introduced what is known as "entropy"; a concept that permeates the field of statistical learning and data science in building and optimizing machine learning models today.

Here is my interpretation and explanation of the concept. Cheers.

https://gradiently.io/a-simple-explanation-of-shannon-entrop...

Hey all. I recently built a github plugin that uses AI to generate a summary of your pull requests automatically. I wanted to share it with you guys here and collect some feedback. Let me know your thoughts, and if you have any suggestions.

Website: https://codetorch.ai/

Github Plugin: https://github.com/apps/codetorch-ai

Documentation: https://codetorch.gitbook.io/codetorch-documentation/#introd...