Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @jmtomczak
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @jmtomczak
-
Jakub Tomczak proslijedio/la je Tweet
Learning Discrete Distributions by Dequantization. http://arxiv.org/abs/2001.11235 pic.twitter.com/ph98xmDp3w
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Learning Discrete Distributions by Dequantization. http://arxiv.org/abs/2001.11235
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Very excited to share "Learning Discrete Distributions by Dequantization" (https://arxiv.org/abs/2001.11235 ) in collaboration with
@TacoCohen and@jmtomczak, from my internship at@Qualcomm. We explore different methods and distributions for dequantization and reach 3.06 bpd on CIFAR10.pic.twitter.com/IGJa7YV8YG
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
In short there are two distributions: the dequantizing distribution q(v|x) and the density model p(v). In this paper we analyse how these distributions interact. Especially take a look at (c), which is the same density model from the previous figure, with a flow as dequantizer.pic.twitter.com/PKDM9dFrMR
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
This is an extension of very cool work by J. Ho et al (https://arxiv.org/abs/1902.00275 ) and
@lucastheis et al (https://arxiv.org/abs/1511.01844 ). Digital media is generally discrete. The likelihood of continuous distributions can be arbitrarily high (a). Instead we add noise to dequantize (b).pic.twitter.com/Cz5qbRSgec
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
My first experience with evolutionary robotics
Learning a neural-network-based controller of a robot with evolvable morphology. A great outcome from a collaboration with @GongjinLan,@matteo_phd, Fuda,@DiederikRo and@gusz_e! The arXiv version: https://arxiv.org/abs/2001.07804 pic.twitter.com/QaHRz4nJNM
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
What better way to start a new year than by applying for a great tenure track position in a great city (Amsterdam) in a great team: AI systems that collaborate with people instead of replacing them: https://workingat.vu.nl/ad/assistant-professor-tenure-track-hybrid-intelligence/5neg8m … (closes in 3 days).
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Really exciting to have my first paper accepted at
@iclr_conf#ICLR2020! It provides the first group theoretical approach towards equivariant visual attention. Nice things coming up next!@erikjbekkers@jmtomczak@Mark. Co-Attentive Equivariant Nets: https://openreview.net/forum?id=r1g6ogrtDr …pic.twitter.com/jppwAD8NJy
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Happy to share that our paper "Batch-shaping for learning conditional channel gated networks" has been accepted at
#ICLR2020#ICLR Paper: https://openreview.net/forum?id=Bke89JBtvB … Joint work with@TiRune &@wellingmaxHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
VAEs enthusiasts, if you are at
#NeurIPS2019 you should come to poster 110 today at 10.45. We'll be presenting BIVA, and show you how to build powerful deep hierarchical VAEs. Paper: https://papers.nips.cc/paper/8882-biva-a-very-deep-hierarchy-of-latent-variables-for-generative-modeling …@larsmaaloe@valentinlievin@OleWinther1Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Curious about how to generate bike trips? Check
@AleZonta presentation "Generation of Human-Like Movements Based on Environmental Features" @ IEEE SSCI 2019 Xiamen#IEEESSCI2019#ComputationalIntelligence#trajectory#generation#humanlikeness#neuralnetworkspic.twitter.com/I03cXlzejn
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Want to understand how to evolve robot predators?
@GongjinLan just presented us how: "Simulated and Real-world Evolution of Predator Robots" @ IEEE SSCI 2019#Xiamen#China#IEEESSCI2019#ComputationalIntelligence#predator#prey#evolutionaryroboticspic.twitter.com/keNfU52kbD
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Check out our extensive review paper on normalizing flows! This paper is the product of years of thinking about flows: it contains everything we know about them, and many new insights. With
@eric_nalisnick,@DeepSpiker,@shakir_za,@balajiln. http://arxiv.org/abs/1912.02762 Thread
https://twitter.com/DeepSpiker/status/1202868429780336640 …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Looking for something to read in your flight to
#NeurIPS2019? Read about Normalizing Flows from our extensive review paper (also with new insights on how to think about and derive new flows) https://arxiv.org/abs/1912.02762 with@gpapamak@eric_nalisnick@DeepSpiker@balajiln@shakir_zapic.twitter.com/EWh8Aui7n0
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Huge congratulations to Patrick Putzky and the rest of our team for winning the “single coil” track! I am very proud!https://twitter.com/facebookai/status/1201161467145465858 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
Postdoc positions available at ETH Zurich Foundations of Data Science initiative:https://math.ethz.ch/sfs/eth-foundations-of-data-science/postdoctoral-positions.html …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
I had a great pleasure to give a talk next to
@rpascanu, M. Jankowiak, J. Sivic,@aosokin_ml, A. Grabska-Garbinska,@fjwolski, M. Andrychowicz, J. Henriques, W. Kotlowski, G. Varol among others at amazing@ML_in_PL about deep generative modeling. Slides: https://jmtomczak.github.io/pdf/MLinPL_2019_11_24.pdf …pic.twitter.com/ZnGbCqrZHZ
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
We have a vacancy for an asst. prof. on Hybrid Intelligence: AI systems that collaborate with people instead of replacing them. We offer a top-100 Dept (Times Higher), a top-20 city (Mercer) and a fully funded PhD student to supervise https://workingat.vu.nl/ad/assistant-professor-tenure-track-hybrid-intelligence/5neg8m … (closes 6 Jan)
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
JOB VACANCY: Interested in doing a Postdoc in Computational Statistics/Statistical Machine Learning in Oxford with me and Chris Holmes, please apply: https://my.corehr.com/pls/uoxrecruit/erq_jobspec_details_form.jobspec?p_id=144065 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jakub Tomczak proslijedio/la je Tweet
.
@facebook czy@Google mogą wysyłać w świat przekaz, że#sztucznainteligencja jest już superzaawansowana. Znawcy tematu widzą jednak, jak bardzo te wszystkie algorytmy są jeszcze ograniczone - o trendach w badaniach#AI mówi@jmtomczak http://www.naukawpolsce.pl/aktualnosci/news%2C79574%2Cinformatyk-o-sztucznej-inteligencji-ona-jeszcze-nie-jest-tak-zaawansowana …#MLinPLHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.