Wrote a little blog post about some subject I am teaching these days. The blog post is about how you can use so called 'image embeddings' to find and cluster similar images.

https://michielbbal.github.io/2024/09/05/Use-embeddings-for-clustering-images-and-finding-similar-images.html

#AI #CV #howcomputerslearn

Use image embeddings for finding similar images and clustering images

By using Deep Learning for images we can create so called ‘image embeddings’. An image embedding, is an image converted to a set of numbers (called a vector) using an AI model. In this blog post, we’ll explore how to use image embeddings for similarity search and clustering, with a focus on OpenAI CLIP, cosine similarity, and KMeans clustering.

Some notes on Applied Artificial Intelligence

This is very nice blog on Compound AI systems, an important step in creating more reliable AI systems.

https://bair.berkeley.edu/blog/2024/02/18/compound-ai-systems/

#AI #howcomputerslearn

The Shift from Models to Compound AI Systems

The BAIR Blog

The Berkeley Artificial Intelligence Research Blog

Keras Core is the latest release of the popular python keras package, that is widely used in deep learning.

In Keras Core you can use Keras, Pytorch,Tensorflow and JAX all in one package.

Blog: https://keras.io/keras_core/announcement/

#howcomputerslearn #keras #pytorch #tensorflow

Keras: Deep Learning for humans

Keras Core documentation

Stanford research has published a document on how 10 Large Language Models are compliant with the new EU AI Act.

It turns out that the open source model BLOOM (https://huggingface.co/bigscience/bloom) adheres best to the AI Act.

BLOOM has been developed by >1000 scientists and is hosted by @huggingface

Source:
https://crfm.stanford.edu/2023/06/15/eu-ai-act.html

#AI #LLM
#OpenSource
#howcomputerslearn
#hoecomputersleren

bigscience/bloom · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

I have created a Jupyter Notebook and compared 10 examples of AI-generated text with 10 examples of human written text. Using python code I calculated and compared their perplexity.

After just testing 20 examples, we can see in the graph below a clear difference between AI-generated and human written text using their perplexity-score.

This is promising, but lot more testing needed.
#howcomputerslearn
#hoecomputersleren

Repo:
https://github.com/MichielBbal/test_ai_detectors

GitHub - MichielBbal/test_ai_detectors: PoC to detect whether text is AI generated or written by a human

PoC to detect whether text is AI generated or written by a human - GitHub - MichielBbal/test_ai_detectors: PoC to detect whether text is AI generated or written by a human

GitHub

This means that a the hundreds of millions spend on developing ChatGPT and Bard might be a waste of money.

Blog: https://simonwillison.net/2023/May/4/no-moat/

Leaked doc: https://www.semianalysis.com/p/google-we-have-no-moat-and-neither

Vicuna: https://lmsys.org/blog/2023-03-30-vicuna/

#howcomputerslearn
#hoecomputersleren

Leaked Google document: “We Have No Moat, And Neither Does OpenAI”

SemiAnalysis published something of a bombshell leaked document this morning: Google “We Have No Moat, And Neither Does OpenAI”. The source of the document is vague: The text below is …