Neural Nets, Transformers & LLMs

Retrieval Augmented Generation

Retrieval Augmented Generation (RAG) for LLM systems clearly explained 👇 pic.twitter.com/2N0pOCC78B
— David Andrés 🤖📈🐍 (@daansan_ml) June 29, 2024

How LLMs work videos

How LLMs work videos on YouTube

Evolution of Deep Learning

Evolution of Deep Learning by Hand ✍️ As my tribute to Geoff Hinton's Nobel Prize, I drew this animation to illustrate the key idea behind Hinton's major contributions to deep learning over the years, with artistic liberty. pic.twitter.com/1SnyIme3wY
— Tom Yeh (@ProfTomYeh) October 9, 2024

Can Large Language Models truly reason?

1/ Can Large Language Models (LLMs) truly reason? Or are they just sophisticated pattern matchers? In our latest preprint, we explore this key question through a large-scale study of both open-source like Llama, Phi, Gemma, and Mistral and leading closed models, including the… pic.twitter.com/yli5q3fKIT
— Mehrdad Farajtabar (@MFarajtabar) October 10, 2024

How LLMs work, clearly explained with visuals

How LLMs work, clearly explained with visuals: pic.twitter.com/7hmVirq6YD
— Sumanth (@Sumanth_077) October 1, 2024

Tokenization — turning text into a sequence of integers — is a key part of generative AI, and most API providers charge per million tokens.

Tokenization — turning text into a sequence of integers — is a key part of generative AI, and most API providers charge per million tokens. How does tokenization work? Learn the details of tokenization and RAG optimization in Retrieval Optimization: From Tokenization to Vector… pic.twitter.com/AsOI8UrXkn
— Andrew Ng (@AndrewYNg) October 2, 2024

Language Language Models are Zero-Shot Reasoners

https://twitter.com/drummatick/status/1841772670821732585

> for ML beginners
> for ML intermediates
> for ML experts
> for anyone who wants to understand why LLMs are so awesome

This paper👇 https://t.co/jiyv9URcxj pic.twitter.com/im0VRzk7OH
— Saurabh Kumar (@drummatick) October 3, 2024

Visualization of what a model like ChatGPT does to produce such efficient responses!

Visualization of what a model like ChatGPT does to produce such efficient responses! pic.twitter.com/CxgPO5JKRC
— ₕₐₘₚₜₒₙ — e/acc (@hamptonism) August 9, 2024

Transformers (LLMs) clearly explained with visuals:

Transformers (LLMs) clearly explained with visuals: pic.twitter.com/GvFcRTCX46
— Shubham Saboo (@Saboo_Shubham_) August 10, 2024

ML Hyperparameters

Ml Hyperparameters pic.twitter.com/lCKolmFsGj
— Aishwarya Nevrekar (@nevrekaraishwa2) August 4, 2024

Convolutional Neural Networks visualisation

Convolutional Neural Networks Visualization
pic.twitter.com/EUD5Dh4Ddz
— ₕₐₘₚₜₒₙ — e/acc (@hamptonism) August 2, 2024

Can LLMs reason?

The question of whether LLMs can reason is, in many ways, the wrong question. The more interesting question is whether they are limited to memorization / interpolative retrieval, or whether they can adapt to novelty beyond what they know. (They can't, at least until you start…
— François Chollet (@fchollet) July 26, 2024

Neural Networks visualised

Neural network visualized! pic.twitter.com/tYiwhUv4JJ
— Daily Dose of Data Science (@DailyDoseOfDS_) July 15, 2024

Single Artificial Neuron in a Neural Network

https://www.linkedin.com/posts/skphd_machinelearning-deeplearning-datascience-activity-7276974765418930176-2Of4

Attention in Transformers, visually explained (by @3blue1brown)

The latest @3blue1brown video on the attention mechanism is a classic story of how just repackaging known information in a new way (often unique to you) can be a huge value add. Great video.

(this is also why sapiens was so successful)https://t.co/0IrtElcjwx
— Nathan Lambert (@natolambert) April 7, 2024