neuralvertigo's profile picture. Applied scientist @AWS AI/ML. Ex @Meta|@Cambridge_Uni|@ucl|@ecentua 🇬🇷.
No cats.

Dionysis Manousakas

@neuralvertigo

Applied scientist @AWS AI/ML. Ex @Meta|@Cambridge_Uni|@ucl|@ecentua 🇬🇷. No cats.

We're hiring Summer Interns! Join our NYC team to work on cutting-edge synthetic data generation for LLMs. This is a unique opportunity to dive into foundation models, and make a real impact. If interested, feel free to share your resume with me! amazon.jobs/en/jobs/279370…


Dionysis Manousakas 已转帖

Another day another launch 😬 🚀🥳 I am super excited to announce that we open sourced Renate, a #ContinualLearning library to automatically retrain and retune #DeepNeuralNetworks. This is joined work with @610v4nn1__ , @lukas_balles and Martin Wistuba! github.com/awslabs/Renate

github.com

GitHub - awslabs/Renate: Library for automatic retraining and continual learning

Library for automatic retraining and continual learning - awslabs/Renate


Dionysis Manousakas 已转帖

A new release of Syne Tune is out! syne-tune.readthedocs.io/en/latest/inde… 🥳 Our goal is to make machine learning more reproducible. The library offers a broad range of asynchronous hyperparameter optimisation algorithms that can be run locally or in the cloud. #hpo #automl #DeepLearning

The (early) Christmas gift you all have been waiting for ;-). New release of Syne Tune! Key changes are a much better documentation, support of SageMaker warm pools and integration of YAHPO. github.com/awslabs/syne-t… Thanks to the team for this great effort!



Dionysis Manousakas 已转帖

The first version of our #hpo chapter for the D2L book is out! d2l.ai/chapter_hyperp… . This is joint work with @cedapprox and Matthias Seeger.


Dionysis Manousakas 已转帖

Time to share our paper “Black-box Coreset Variational Inference” accepted to #NeurIPS2022 with @neuralvertigo and Hippolyt Ritter. We re-derive Bayesian core sets under an approximate inference lens to make them generally applicable. Paper link arxiv.org/abs/2211.02377 1/n


Dionysis Manousakas 已转帖

Diffusion models like #DALLE and #StableDiffusion are state of the art for image generation, yet our understanding of them is in its infancy. This thread introduces the basics of how diffusion models work, how we understand them, and why I think this understanding is broken.🧵

tomgoldsteincs's tweet image. Diffusion models like #DALLE and #StableDiffusion are state of the art for image generation, yet our understanding of them is in its infancy. This thread introduces the basics of how diffusion models work, how we understand them, and why I think this understanding is broken.🧵

Dionysis Manousakas 已转帖

As many Ukrainian IT engineers right now I was left without a job 😔I am a Data Scientist with 5 years of experience in full-cycle ML models development (including deployment) and in managing data science teams. If anyone sees job positions I could fit, ping me 🙏❤️

dark_matter88_'s tweet image. As many Ukrainian IT engineers right now I was left without a job 😔I am a Data Scientist with 5 years of experience in full-cycle ML models development (including deployment) and in managing data science teams. If anyone sees job positions I could fit, ping me 🙏❤️

Dionysis Manousakas 已转帖

Give, I say, all your strength to your work, make it your total concern. And don’t forget your work even in times of trial or when you near your end.


Dionysis Manousakas 已转帖

Good times! Welcome to two new members on the team, Dionysis Manousakas @neuralvertigo and Hippolyt Ritter, who will be working with me on the interface of deep learning and probabilistic inference. Very excited to work with both of you!


Dionysis Manousakas 已转帖

He who hopes to grow in spirit will have to free himself from obedience and respect.


Dionysis Manousakas 已转帖
fhuszar's tweet image.
fhuszar's tweet image.
fhuszar's tweet image.

Dionysis Manousakas 已转帖

The Call for Papers for #AISTATS2020 is out at aistats.org The submission deadline is 8 October 2019. The conference will be held in Palermo, Italy, from 3 to 5 June 2020.


Dionysis Manousakas 已转帖

Mathematicians Caucher Birkar, Alessio Figalli, Peter Scholze and Akshay Venkatesh have just been awarded the Fields Medal. Computer scientist Constantinos Daskalakis won the Nevanlinna Prize. Read our exclusive, in-depth profiles of all five winners. buff.ly/2OzGetJ

QuantaMagazine's tweet image. Mathematicians Caucher Birkar, Alessio Figalli, Peter Scholze and Akshay Venkatesh have just been awarded the Fields Medal. Computer scientist Constantinos Daskalakis won the Nevanlinna Prize. Read our exclusive, in-depth profiles of all five winners. buff.ly/2OzGetJ

Dionysis Manousakas 已转帖

Judea Pearl claims all we do in ML is curve fitting. I wrote this post to explain that claim and introduce the basics of causal inference to ML folks. Machine Learning beyond Curve Fitting: An Intro to Causal Inference and do-Calculus inference.vc/p/f01e8d56-511…

inference.vc

ML beyond Curve Fitting: An Intro to Causal Inference and do-Calculus

Since writing this post back in 2018, I have extended this to a 4-part series on causal inference: * ➡️️ Part 1: Intro to causal inference and do-calculus [https://www.inference.vc/untitled] * Part...


Loading...

Something went wrong.


Something went wrong.