About Me — Andrei

A mind inside a body. Probably the same as you, but with a few differences.

I was there once (Image by the author)

If we think about it, it’s a miracle when two minds meet each other, from so many possibilities among all the people who are alive now and from all the people who ever existed. We could have met Napoleon, Caesar, or David, but in the end, we meet each other, here, now.
Meeting you here is not as good as meeting you in reality. I mean, the other reality, the real reality, so let’s pretend that we meet in that one.

If we had met in person, the conversation would probably have started as follows:

At the beginning, we would both say: ‘Hi!’

I will say: ‘Hi! I am Andrei.(Now you know that I am one of the many Andreis that existed on this earth. I start to get a personality.)

Then, just to give a bit of momentum to the conversation, probably we would ask each other:

Where are you from?

I would say that I am from Romania. You will probably make a joke about Dracula and Transylvania. I will encourage you to visit the country. I’ll tell you that it’s gorgeous; we have bears on the roads, delicious food good for your soul and bad for your liver. While I would tell you this, you will see how my pupils get larger and how my nose starts to sniff the air by itself, as if it’s already smelling the vapor from the fried sausages. I will stop here with this. I don’t want to torture myself, and you don’t have time for this now. It’s my time to ask you, ‘Where are you from?’

And after you tell your story, you will ask me:

And what brought you here? What is your plan?

I plan to write here about everything that I would like to know, and don’t have time to explore alone. I know it is selfish, but it is what it is. I hope it will work. I will write here about my ideas, and you and hopefully a bunch of other people will comment, criticize, and point out the flaws in my way of thinking.

But in general, what ideas would you have?
I can tell you what I would start with, but who knows where the wind of curiosity will push me in this adventure. Now, I would like to think about how we can recreate human behaviours inside a neural network and probably will approach topics like:

1. How can we induce reflexes inside an ML model?
2. We do not need to stop when we learn, for training and inference happen at the same time. How can we do that?
3. A neuron is a powerful tool, but the power of a hormone should not be underestimated. What is the equivalent of a hormone in an ML model?
4. How can we create a model that adapts its dimensions while learning? How can we encourage it to create new connections, but at the same time remain as efficient as possible?

These are just a few of the things that my mind wishes to explore. What would you like to think about together with me?

At this point you are probably curious and may ask me:

Okay, interesting, but how did you get into this?

I should have been a physicist, but I realized that what I enjoy is not physics but solving interesting problems that have an application.

I was lucky enough to jump into the quantum computer train before it started to turn into a rocket (although whether it does become one remains to be seen), in the end, I got to apply quantum-inspired algorithms to LLM, and design ML algorithms for creating new materials, which is what I hope to succeed at in the years of my PhD.

Working at this convergence of fields made me ask questions, but as any scientist knows, at one moment you will have a drawer full of questions covered by dust, and this is my attempt to brush the dust off my pile of questions.

Now you know almost everything that I want you to know about me.But you still don’t know me, and it’s okay; we know enough about each other to function together.

It’s not necessary, but if you wish to know more, a good picture of my professional history can also be found here:

— LinkedIn page: https://www.linkedin.com/in/voicu-tomut/

Please let me know your ideas, and if you can recommend me some documentation, do it!