hyle's profile picture. Computer Scientist

Andrea Baresi

@hyle

Computer Scientist

Convert EPUBs to legacy Kindle format without the bloat. Minimal python, standard library only, single file. When you need to read something without installing Calibre github.com/hyle/epub2mobi…


Where 'AGI fully realized' will, for a while, just be current LLMs scaled a bit more, with fewer hallucinations, better tooling, and long-term memory.

AGI fully realized will actually give people a choice: relax or work harder on bigger more ambitious things than you ever thought possible



From "walled gardens" to "agentic walled gardens". The platform still controls which counterparties you can reach, the protocols, the ranking, fees, and even which "service agents" are visible. Walled gardens aren’t automatically bad, but... #AgenticEconomy


Culture influences institutions, and institutions reinforce or reshape culture, but the two are not the same. You can change an institution with the stroke of a pen; changing culture requires changing minds.


Technology is a lever, not a compass.


Adopting some of the latest coding tools puts us in a kind of Turing run-around-the-house (very serious) game for programming--where we can instruct an agent to handle (a very proficiently instructed portion of) our coding chores while we stay fit.


Andrea Baresi reposted

This year’s chemistry laureates Demis Hassabis and John Jumper have developed an AI model, AlphaFold2, to solve a 50-year-old problem: predicting proteins’ complex structures. Check out two examples of protein structures determined using AlphaFold2. First up, a bacterial enzyme…


Andrea Baresi reposted

BREAKING NEWS The Royal Swedish Academy of Sciences has decided to award the 2024 #NobelPrize in Physics to John J. Hopfield and Geoffrey E. Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks.”

NobelPrize's tweet image. BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2024 #NobelPrize in Physics to John J. Hopfield and Geoffrey E. Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks.”

Il vero genio non lavora sulle alte probabilità. #LLMs


Andrea Baresi reposted

What's next for AlphaFold and the AI protein-folding revolution go.nature.com/3vq8YeC


Andrea Baresi reposted

Real-time person segmentation in your browser with TensorFlow.js

BodyPix 2.0 has been released, including multi-person segmentation support and a new live demo! To learn more, read the post by @tylerzhu3, @oveddan, @greenbeandou, @dsmilkov, @karlssonper, @ire_alva, @nsthorat. Details here → goo.gle/bodypix2



Andrea Baresi reposted

Using AI to help scientists solve big questions is at the core of DeepMind’s mission. Today we’re delighted to announce our first significant milestone: a successful application of machine learning to the protein folding problem. deepmind.com/blog/alphafold/ #AlphaFold (1/5)


Andrea Baresi reposted

Bigger models are better models so we built GPipe to enable training of large models. Results: 84.3% on ImageNet with AmoebaNet (big jump from other state-of-art models) and 99% on CIFAR-10 with transfer learning. Link: arxiv.org/abs/1811.06965

quocleix's tweet image. Bigger models are better models so we built GPipe to enable training of large models. Results: 84.3% on ImageNet with AmoebaNet (big jump from other state-of-art models) and 99% on CIFAR-10 with transfer learning. Link: arxiv.org/abs/1811.06965

Andrea Baresi reposted

Excited to share what I've been working on these last few months! Spinning Up in Deep RL, an educational resource that we hope can bring down the barrier to entry in RL. spinningup.openai.com/en/latest/ blog.openai.com/spinning-up-in…


Andrea Baresi reposted

The MAME RL Algorithm Training Toolkit: A Python library that can provide a Gym-like API around almost any old arcade game. They show how to set up new ROMS, and provide RL example for “Street Fighter III Third Strike: Fight for the Future (Japan version)” github.com/M-J-Murray/MAM…


Andrea Baresi reposted

Here is the same dynamic RNN implemented in 4 different frameworks (TensorFlow/Keras, MXNet/Gluon, Chainer, PyTorch). Can you tell which is which?

fchollet's tweet image. Here is the same dynamic RNN implemented in 4 different frameworks (TensorFlow/Keras, MXNet/Gluon, Chainer, PyTorch). Can you tell which is which?

Andrea Baresi reposted

Q-map: a Convolutional Approach for Goal-Oriented RL. We use a convolutional autoencoder-like architecture to quickly learn to reach on-screen coordinates. ArXiv: arxiv.org/abs/1810.02927 Videos and code: sites.google.com/view/q-map-rl


Loading...

Something went wrong.


Something went wrong.