Back to the CLIP neural network that converts text to images and VQGAN!
This time, I wrote a code that processed many prompts and extracted a bunch of tags to a graph.
There it a visualization I got:
A list of CLIP-based image generators to use and learn from.
Katherine Crowson; VQGAN and CLIP, z+quantize:
Katherine Crowson; VQGAN and CLIP, codebook sampling method:
Vadim Epstein; SIREN and CLIP:
There's a source URL for a C example interacting with Speech-dispatcher service, because I definitely did not shoot myself in the foot while coding this tiny script and will never ever need to read it later to fix something.
Got tired while reading articles.
Made a tiny thing to send text from Chrome to RHVoice. Let's call it "Something TTS".
On one side we have a REST server, on the other side we have a Chrome plugin that adds an item into a right-click menu.
Typically I would ask myself why bother, but Chrome plugins that are popular did not work for me / interrupted while reading.
(It's silly, I know.)
Just heard that pleochroism is a thing. And it's a beautiful thing.
(Picture is not directly related)
plot takes in data and plots it using line characters. plot automatically scales the graph, allowing for sudden, unexpected changes to be seen clearly and immediately. plot can display many sources of live data using pipelines. plot can also display the data using a different character set than the default ASCII or Unicode.
Website 🔗️: https://github.com/annacrombie/plot
Sometimes it's hard to explain the homunculus problem / argument.
(Yes, there's no single part of the brain doing the thinking, because with that rhetoric a part of this part will soon do the thinking, so on and so forth.)
From now on, this webcomic page will make my life easier: https://falseknees.com/247.html
The previous explanation from "Computational Cognitive Neuroscience" about memory de-indexing sheds some light on that: the article is about the same brain regions: CA1, CA3 and DG.
Full size text:
An article "A Hippocampal Cognitive Prosthesis: Multi-Input, Multi-Output Nonlinear Modeling and VLSI Implementation" by Theodore W. Berger, et al. (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3395724/) tells about a brain implant with 16 analog-to-digital converters and 32 channels of output.
I've been discussing it with Prozion (https://github.com/prozion) before and he has a very good question."How does the implant function at all with that amount of channels?"
A mental note: "Computational Cognitive Neuroscience" videos have a very informative moment about the inner workings of the hyppocampus in a context of the memory retrieval, where Randall O'Reilly compares data retrieval to a hash:
CA1 region of the hyppocampus as a "memory de-indexer", unpacking activity pattern when we retrieve the memory and activating related cortical areas.
Dentate gyrus and CA3 as networks forming the said pointer.
There's a blog / newsletter about AI by Jack Clark, co-founder of Anthropic, previously a policy director of OpenAI: IMPORT AI.
It might be rather interesting for people playing with Transformers and NLP in general; Anthropic plans to do quite a bit:
A tiny note for myself about "neurons that fire together, wire together": there's a temporal precendence, one cell should fire before the other so the learning, the actual wiring, can happen.
Listening to "CCN Course 2020, Neuron 11: Ions". It tells about membrane potential, sodium and potassium pumps, allowing our neurons to work.
…Have you ever thought that a neuron had evolved in a sea water?
We carry a trace of an ancient ocean with us throughout our lives.
(Using a Silas Baisch image from Unsplash as a base for CLIP render.)
Continuing the topic of "how our brains work", both the Neuronify program and "Handbook of Brain Microcircuits" book by Gordon M. Shepherd and Sten Grillner are very relevant.
We have many types of neuron cells arranged in different ways to continue functioning.
Some of them help to control our heart, some help with detecting motions or some other type of input information.
The fact remains, though: some neuron arrangements are useful for specific groups tasks.
Great news! EleutherAI's GPT-J-6B, a 6 billion parameter model trained on the Pile dataset is released.
It is kinda similar to OpenAI's curie API. It is, afaik, Apache-licensed. In any case, have a look!
Web Demo: https://6b.eleuther.ai/
Blog post: https://arankomatsuzaki.wordpress.com/2021/06/04/gpt-j/
Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.