curiosity   2765

« earlier    

Reinforcement Learning with Prediction-Based Rewards
We’ve developed Random Network Distillation (RND), a prediction-based method for encouraging reinforcement learning agents to explore their environments through curiosity, which for the first time1 exceeds average human performance on Montezuma’s Revenge. RND achieves state-of-the-art performance, periodically finds all 24 rooms and solves the first level without using demonstrations or having access to the underlying state of the game.

RND incentivizes visiting unfamiliar states by measuring how hard it is to predict the output of a fixed random neural network on visited states. In unfamiliar states it’s hard to guess the output, and hence the reward is high. It can be applied to any reinforcement learning algorithm, is simple to implement and efficient to scale. Below we release a reference implementation of RND that can reproduce the results from our paper.
machine-learning  reinforcement-learning  curiosity  the-mangle-in-practice  to-write-about  also-note-presentation 
5 days ago by Vaguery
ytCropper - Crop YouTube Videos Online
I'm always looking for new and simpler ways to curate moments of Ss #curiosity by strategically cropping @YouTube vidoes. yTCropper is my new favorite tool! #edtech #spark3
IFTTT  Diigo  curiosity  edtech  spark3  fav_tweet 
13 days ago by occam98
Towards a neuroscience of active sampling and curiosity | Nature Reviews Neuroscience
> "...we distinguish between information sampling, in which organisms reduce uncertainty relevant to a familiar task, and information search, in which they investigate in an open-ended fashion to discover new tasks"
27 days ago by tonyyet
How to Talk to People, According to Terry Gross
Nov. 17, 2018 | The New York Times | By Jolie Kerr.

(1) “Tell me about yourself,” a.k.a the only icebreaker you’ll ever need.
(2) The secret to being a good conversationalist? Curiosity.
(3) Be funny (if you can). “A good conversationalist is somebody who is fun to talk to,” she said. Ms. Gross, it’s worth noting, is very funny. If you can’t be funny, being mentally organized, reasonably concise and energetic will go a long way in impressing people.
(4) Preparation is key. “It helps to organize your thoughts beforehand by thinking about the things you expect you’ll be asked and then reflecting on how you might answer,” think through where your boundaries are, so that you’re not paralyzed agonizing over whether you’re willing to confide something or not.”

In a job interview, organizing your thoughts by thinking about the things you expect you’ll be asked and reflecting on how you might answer can help you navigate if things start to go badly.
(5) Take control by pivoting to something you want to talk about.
(6) Ms. Gross doesn’t want you to dodge questions. But if you’re going to, here’s how: Say, “I don’t want to answer that,” or, if that’s too blunt, hedge with a statement like, “I’m having a difficult time thinking of a specific answer to that.” Going the martyr route with something like, “I’m afraid by answering that I’m going to hurt somebody’s feelings and I don’t want to do that,” is another option.
(7) Terry pays attention to body language. Be like Terry.
(8) When to push back, and when not to.
body_language  Communicating_&_Connecting  conversations  curiosity  howto  humour  preparation  tips  nonverbal  posture  ice-breakers 
28 days ago by jerryking
Two Hundred Fifty Things an Architect Should Know — R / D
An archive of critical writing about design. From architecture and urbanism
to product, fashion, graphic and beyond.
design  writing  architecture  építészet  curiosity  lists  250  architect  article  bestof 
29 days ago by atran
I was Pat Tillman’s wife, but I can’t speak for him. Neither can you. - The Washington Post
Quote: "... to err on the side of passion is human and right and the only way I’ll live."
nfl  life  passion  curiosity  reading 
4 weeks ago by ajohnson1200
Differentiable Monte Carlo Ray Tracing through Edge Sampling
Gradient-based methods are becoming increasingly important for computer graphics, machine learning, and computer vision. The ability to compute gradients is crucial to optimization, inverse problems, and deep learning. In rendering, the gradient is required with respect to variables such as camera parameters, light sources, scene geometry, or material appearance. However, computing the gradient of rendering is challenging because the rendering integral includes visibility terms that are not differentiable. Previous work on differentiable rendering has focused on approximate solutions. They often do not handle secondary effects such as shadows or global illumination, or they do not provide the gradient with respect to variables other than pixel coordinates.

We introduce a general-purpose differentiable ray tracer, which, to our knowledge, is the first comprehensive solution that is able to compute derivatives of scalar functions over a rendered image with respect to arbitrary scene parameters such as camera pose, scene geometry, materials, and lighting parameters. The key to our method is a novel edge sampling algorithm that directly samples the Dirac delta functions introduced by the derivatives of the discontinuous integrand. We also develop efficient importance sampling methods based on spatial hierarchies. Our method can generate gradients in times running from seconds to minutes depending on scene complexity and desired precision.
graphics  programming  curiosity 
4 weeks ago by naijeru
Key Papers in Deep RL — Spinning Up documentation
What follows is a list of papers in deep RL that are worth reading. This is far from comprehensive, but should provide a useful starting point for someone looking to do research in the field.
AI  deep_learning  programming  curiosity 
5 weeks ago by naijeru
"Cosmopolitanism is not an elite identity; it is an attitude of about what lies beyond the boundaries of…
curiosity  from twitter_favs
5 weeks ago by dermotcasey

« earlier