• LLMs are a kind of neural network
• They are trained by massive data sets
( Trillions of tokens )
• LLMs learn about patterns and language from the massive text that they were trained on
• LLMs are like a very complex autocomplete
• Do not believe LLMs’ output without verifying
• Zero shot learning : a prompt without no particular example provided
• Few shot learning : a prompt that contains a few examples

8 things you should know about LLMs by arxiv :

1. LLMs predictably get more capable with increasing investment, even without targeted innovation.

2. Many important LLM behaviors emerge unpredictably as a byproduct of increasing investment.

3. LLMs often appear to learn and use representations of the outside world.

4. There are no reliable techniques for steering the behavior of LLMs.

5. Experts are not yet able to interpret the inner workings of LLMs.

6. Human performance on a task isn't an upper bound on LLM performance.

7. LLMs need not express the values of their creators nor the values encoded in web text.

8. Brief interactions with LLMs are often misleading.

for summarizing youtube videos NoteGPT is really useful. it provides free accurate summarization and mindmap

this is the mindmap that noteGPT generated for the video above

Straight forward tutorial. gotta watch it dozens of times

Let's talk about GAN and Fine-Tuning

Two topics I'm excited to talk about

Starting with GAN

go look at this website

a really fascinating example of GAN

this site generates superrealistic human faces in seconds

and no

it does not have a database and with each refresh it generates a complete new face

using GAN

Generative Adversarial Networks

what is it then?