Reinforcement Learning Study Group
Hey everyone,
my name is Kay. I’m new to the forum and I came here with a specific goal in mind:
I’m putting together a crew for a reinforcement learning study group.
Main objectives:
Mathematical Foundations: We will work through key passages in Sutton & Barto’s Book to get a good foundation of RL
Research Papers: We will follow the “Spinning up in Deep Reinforcement Learning” by OpenAI resource to select key research papers in RL to read.
Code: The above resource can also be used to explore main RL algorithms and ideas which we’ll aim to program from scratch in Pytorch/Tensorflow
Get Practical Skills: In the long term, the goal of this group is to prepare ourselves for work in AI Alignment and adjacent fields as engineers, researchers etc. (Work at DeepMind, OpenAI, MIRI, FHI, etc.)
Important Self-Selection:
It is important for me to be surrounded by ambitious and self-motivated people who are reliable, friendly, and helpful. I expect members of the study group to invest a good amount of time to study and even more to try to teach one another. Members who only plan on participating passively will be asked to become more active or leave the group.
Think of the study group as a sports team. The goal is to put the best people together on the field. Those who aren’t ready to play will have to sit on the sideline.
Other helpful prerequisites are being able to program in Python 3, some knowledge of Probability and Statistics, as well as Linear Algebra and Calculus. In case this seems a little intimidating, don’t worry, err on the side of joining the group. You’ll pretty soon discover whether you’re a good fit or not. Remember, we’re all beginners who are united by the goal of becoming better.
Becoming our Coach:
In the above spirit, I’m also looking for a more advanced coach, someone who knows the terrain of reinforcement learning relatively well and who’s willing to guide the teams’ efforts towards a more fruitful, albeit challenging learning experience.
Anyone who’s familiar with the main concepts in Sutton & Barto’s Book would fit this role.
Schedule:
The study group will start as soon as possible in January. Once there is at least one other participant aside from me, we will find a suitable timetable and meet at least once a week to do pair-programming, code reviews, work through mathematical formulas, discuss papers, and more.
How to Join:
Contact me via kozaronek@gmail.com and introduce yourself, I’m happy to hear from you and excited to get started!
I would like to say that there’s a study group being formed in the AI Alignment Slack server with similar intentions! If you are not a part of that server and would like to join, feel free to email me at melembroucarlitos@gmail.com telling me a bit about yourself and your hopes and intentions and I’ll send you an invite.
Alternatively you may want to join here: https://join.slack.com/t/ai-alignment/shared_invite/zt-fkgwbd2b-kK50z~BbVclOZMM9UP44gw
Thanks, Pablo. This invite worked. Good to know that there’s already such a big community.
Thanks for sharing Lucas, I’ll shoot you a message.
What is your part in all of this, are you a learner too?
What’s the weekly time commitment of this study group?
I expect a time commitment of about 10 hours, but there’s room for more.
Most probably I’ll be spending about 20 hours per week or more, so if you’re willing to do more, at the very least there’s one person who’ll be up for the extra mile.
I hope the spinning up content still mostly works—a bunch of the relevant libraries have updated in the intervening years.
I would also recommend the older, but still valuable lecture series on RL by David Silver: https://deepmind.com/learning-resources/-introduction-reinforcement-learning-david-silver
I advise using JAX instead of Tensorflow.