OpenAI
Opening & Intro to RL, Part 1, by Joshua Achiam at 25:11
Intro to RL, Part 2, by Joshua Achiam at 1:48:42
Learning Dexterity, by Matthias Plappert at 2:26:26
AI Safety: An Introduction, by Dario Amodei at 2:58:00
Source
OpenAI – Spinning Up in Deep RL Workshop
I needed for this to come out like two months ago.
Great stuff.
Github Repository link https://github.com/openai/spinningup-workshop
Human preferences is a nice way to learn rewards, but it does not directly address safety. How can we be certain the learned reward acts well across all states? https://twitter.com/robinc/status/1092227059374424064
SpinningUp repo in PyTorch https://github.com/mksmsrkn/spinningup_pytorch [NOT OFFICIAL]
Thanks OpenAI for the great workshop ?
Even if verbose, a conversational translation of the formulae, may increase the accessibility of the information.
A "conversational compiler" for code, including math, might be one of the most important programs ever written.
Thanks for making this available!
Is this beginner friendly? I have experience with DL in supervised learning. But neglible knowledge in RL. Any prerequisites before viewing this talk? Thanks
"Spinning Up is currently only supported on Linux and OSX."
Im intrested why you dont support windows? I mean most of people using windows like me. 🙂
has openai 5 been abandoned? why is there no dota 2 matches between pro and open ai 5 available from last 5 months?
What does one think of these?…might want to look into this terrorist incident regarding AI and having your wives and girlfriends raped over a nanonet, using AI.
How will AI move forward when these are becoming a rampant perfidious threat?
Quick aside question but it does pertain to if I wanted to get nanotech removed for an incidence claim where would be the best place to have this done per someone's suggestion? Should I go about contacting major schools to describe this? I need to get this out into the media…
https://twitter.com/ThePentagram7
figured I'd post this here and see if I can get a bite, bare in mind, I have been harassed by these people nearly nonstop 24hrs a day, 7 days a week. I was having to somewhat simultaneously perform an on-the-spot-interrogation because of how I was treated and that was my only makeshift defense. It may initially come off as preposterous, but I assure you, I wouldn't put my face out there like this if I was joking with a novel token.
Thanks for hosting this workshop. Is there a place I could subscribe to get notifications of upcoming events of this kind?
@2:03:35 it says van Hesselt et al. Pretty sure it is van Hasselt as he was my supervisor 😉
oh god, get a pop filter
He do I put openAi into my homemade game?
1:00:40 guy starts talking about mathematics, cameraman falls asleep
Do counter strike AI what they have 300 ms and pointer vibrations
I want to study machine learning.
-What about his comment on pytorch and tenserflow at 46:46 ? Could someone advise me?
Can you fix the sound ? Ty
This presentation left me feeling like I am not smart enough to learn RL. Now I am bummed.
Thanks for this! Especially Joshua for his explanations of the math. really helped me get an intuition for the main algorithms, thanks!
For the lazy web, here's the link to OpenAI open source repo that has the PDFs for these slides: https://github.com/openai/spinningup-workshop