Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @neil_uai
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @neil_uai
-
Prikvačeni tweet
https://xiaohuzhu.xyz/alignment-newsletter-zh/ … Keep people aligned on AGI/AI all over the world.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
On challenges of learning human preferences...https://twitter.com/StanfordHAI/status/1223259702915686403 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
https://arxiv.org/abs/2001.04413 Cool theory paper presenting a problem that: - can be efficiently learned by SGD with a DenseNet with x^2 nonlin, - cannot be efficiently learned by any kernel method, including NTK.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
How can we predict and control the collective behaviour of artificial agents? Classical game theory isn't much help when there are >2 agents. In our
@iclr_conf paper, we find markets impose useful structure on interactions between gradient-based learners: https://arxiv.org/abs/2001.04678 pic.twitter.com/IeLMcb9f2z
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Is deep learning is actually performing DEEP learning? We may have given the first proof that neural network is capable of efficient hierarchical learning, while existing theory only shows that deep learning can "simulate" non-hierarchical algorithmshttps://twitter.com/MSFTResearch/status/1217198142229102593 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Enough people asked about Coq vs Lean and setoids vs quotients that I tried to summarize what I know:https://gist.github.com/Blaisorblade/78441b030a7c0e20ab1242c8e6662db6 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Studying or teaching Complex Analysis? This online book might be useful.
Complex Analysis: A Visual and Interactive Introduction
https://complex-analysis.com/
#complex#analysis#OpenSource Thanks everyone for your suggestions and reporting issues to improve this project.pic.twitter.com/ElaWs4XjIVHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
I think it is clear that no AI we have at present can create life supporting systems on Mars, nor create new knowledge about cancer or black holes. Much work lies ahead and everyone seriously working on AI knows this fact. Let’s get down to work! Time is ticking.https://twitter.com/yudapearl/status/1215174538087948288 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Big shout out to the wildlife rescuers in Australia (pic: David Mariuz)pic.twitter.com/8gR8cKUDPQ
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
The DotA paper is out!!!https://twitter.com/OpenAI/status/1205541134925099008 …
0:11Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Self-knowledge is the most valuable kind of knowledge, because all knowledge must be filtered through the self. If you don't understand the instrument, you don't understand the measurement.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Oldies but goldies: Frank Rosenblatt, The Perceptron-a perceiving and recognizing automaton, 1957. The first appearance of a neural network (single layer), later generalized to multiple layers. https://en.wikipedia.org/wiki/Perceptron pic.twitter.com/Ysm1xdos2R
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
At OpenAI, we're excited by work in learned optimizers such as Learning To Learn By Gradient Descent By Gradient Descent. These methods will improve the efficiency of our models, and allow for higher quality science. Looking for a founding team member!https://tinyurl.com/tjz8ct3
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Many grand AI challenges remain. If you want to join our mission alongside me,
@lilaibrahim,@koraykv,@ShaneLegg & others, check out our scholarships & open roles. https://deepmind.com/blog/article/strengthening-the-ai-community … 3/3Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
This is a great extension of Coin Run, really curious to see what research it will inspire!https://twitter.com/OpenAI/status/1201909508009627649 …
0:29Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
The max and soft-max (log-sum-exp) of convex functions is convex. Countless applications, from convex duality to exponential families. https://en.wikipedia.org/wiki/LogSumExp pic.twitter.com/pdPvqAHitd
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
This is a resurrection plant. It’s a desert plant that can survive almost complete desiccation. During the dry weather it curls into a tight ball, uncurling when exposed to moisturepic.twitter.com/QSLoHvKBJ4
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Postdoc positions in theoretical machine learning at Princeton CS Dept. Relevant faculty include Elad Hazan, Ryan Adams, Yoram Singer, and me. Mention in cover letter which faculty you are interested in. Best to apply by Dec 15; latest by Jan 10. https://mltheory.cs.princeton.edu/positions/
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Fantastic analogy from
@geoffreyhinton deconstructing the strange but popular school of thought that because deep learning isn't perfect the solution must be to combine it with symbolic AI.https://twitter.com/tabithagold/status/1071189769499996161 …
0:28Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
A mean field evolution is the large number limit of particles systems. The simplest case is gradient descent, which corresponds to an advection equation in the limit. https://en.wikipedia.org/wiki/Advection https://en.wikipedia.org/wiki/Mean_field_theory …pic.twitter.com/1yO5Ti7Tfj
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Xiaohu ZHU proslijedio/la je Tweet
Birkhoff’s contraction for Hilbert’s metric is a key tool to quantify convergence of Markov-Chains (Perron-Frobenius). https://en.wikipedia.org/wiki/Hilbert_metric …pic.twitter.com/9w5CelmTYF
Hvala. 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.