alex peysakhovich
Behavioral economist. AI Researcher at Facebook. MythBusters axiom: “The only difference between screwing around and science is writing it down.”
Last synced from Twitter 14m ago
PyTorch
An open source deep learning platform that provides a seamless path from research prototyping to production deployment.
Discovering Context Effects from Raw Choice Data
Many applications in preference learning assume that decisions come from the maximization of a stable utility function. Yet a large experimental literature shows that individual choices and judgeme...
Wordbank: An open database of children's vocabulary development
2
alex peysakhovich
Really awesome interactive cross-cultural dataset on human language learning. Super fun to play with!
28d
Object Discovery with a Copy-Pasting GAN
We tackle the problem of object discovery, where objects are segmented for a
given input image, and the system is trained without using any direct
supervision whatsoever. A novel copy-pasting GAN framework is proposed, where
the generator learns to...
Judgments of Effort for Magical Violations of Intuitive Physics
People spend much of their time in imaginary worlds, and have beliefs about the events that are likely in those worlds, and the laws that govern them. Such beliefs are likely affected by people’s intuitive theories of the real world. In two...
Judgments of effort for magical violations of intuitive physics
People spend much of their time in imaginary worlds, and have beliefs about the events that are likely in those worlds, and the laws that govern them. Such beliefs are likely affected by people’s intuitive theories of the real world. In three...
The Illustrated Transformer
In the previous post, we looked at Attention – a ubiquitous method in modern deep learning models. Attention is a concept that helped improve the performance of neural machine translation applications. In this post, we will look at The Transformer –...
7 Ways to Age Well in 2019
How to increase your chances of living longer, learn the secret to aging well or just look younger.
alex peysakhovich
@ProfEmilyOster - can you please write a book about health next? I want to know where the research is and I know I can trust you not to write things like this! (from nytimes.com/2018/12/24/wel…) pic.twitter.com/DF6ALckM6A
54d
Robust Multi-agent Counterfactual Prediction
We consider the problem of using logged data to make predictions about what
would happen if we changed the `rules of the game' in a multi-agent system.
This task is difficult because in many cases we observe actions individuals
take but not their...
2
alex peysakhovich
@ChrKroer, Adam Lerer, and myself recently got to asking: how can we make better counterfactual predictions for multi-agent systems? We posted this paper a bit ago with a potential answer arxiv.org/abs/1904.02235 1/4
55d
Postdoc Behavioral Experiments
Post-doctoral fellow at Susan Athey’s lab Susan Athey, the Economics of Technology Professor at Stanford University, has an opening for a 2 year position as a post-doctoral fellow, potentially working as part of her Shared Prosperity and Innovation...
Woman's perfect Tinder date with 'The One' shattered by five simple words
They had been messaging for a while and she thought she knew him pretty well - but she was very, very wrong
A/B Testing by Eduardo M. Azevedo, Alex Deng , Jose Montiel Olea, Justin M. Rao, E. Glen Weyl :: SSRN
Large and thus statistically powerful A/B tests are increasingly popular in business and policy to evaluate potential innovations. We study how to optimally use
2
alex peysakhovich
@jim_savage_ @Susan_Athey I think this is a bad heuristic. The distribution of treatment effects is often fat tailed (see my paper with @deaneckles arxiv.org/abs/1701.01140 or Azevedo et al. papers.ssrn.com/sol3/papers.cf…). The distribution of het effects is also likely fat tailed. 1/n
81d
PyTorch-BigGraph: Faster embeddings of large graphs
Facebook AI has created and is now open-sourcing PyTorch-BigGraph (PBG), a tool that makes it much faster and easier to produce graph embeddings for extremely large graphs.
For cramped New York, an expanding dining scene
Can New York become a player in the country's dining scene? It may be on its way.