Shoan :)

Would love to work with your team on this! Along with experience in agentic systems i also am experienced in RL. Let me know if there is any fit with your team: shoan-raj.github.io/uploads/Shoa...

Congratulations Lucas! I've been wanting to learn more about MORL, any resources you recommend?

Hey man, just saw your projects and wanted to let you know that they seem amazing! Excited to see what you make with RL

This felt like a perfect read to get started with MARL, concise enough to finish it quickly but still having a good depth in the concepts

A First Introduction to Cooperative Multi-Agent Reinforcement Learning Multi-agent reinforcement learning (MARL) has exploded in popularity in recent years. While numerous approaches have been developed, they can be broadly categorized into three main types: centralized ...

I have a draft of my introduction to cooperative multi-agent reinforcement learning on arxiv. Check it out and let me know any feedback you have. The plan is to polish and extend the material into a more comprehensive text with Frans Oliehoek.

arxiv.org/abs/2405.06161

"Code Structure" section explaining structure of the codebase (file by file explanations) in the readme.md of a repo

I wish more repos had this

In many benchmark environments, 10-20 lines of static instructions could leap past square 1 —not perfect, but better than nothing. This makes me think RL excels at refining good systems into great ones rather than starting from a blank slate, which would explain its increasing usage in LLM alignment

Another point I'd like to add is to have a good file structure - having worked with multiple large codebases I'd say this is probably the most useful thing. Once you start doing this properly, modularity within code follows automatically making it much more readable

Campus cats are the best

Do undergrads count? Would love to be added :)

Figure 1: Learning from machine-unique knowledge. Shows that the goal of the paper is extracting new concepts from machines that humans don't know about yet.

Sharing this awesome paper which shows that:
1. You can extract concepts unknown to humans from superhuman agents
2. Those concepts can then seemingly be taught to experts via examples
arxiv.org/abs/2310.16410

The amount of new things I discovered just by scrolling on this app is insane

Lovely username btw

The best part of social media is paper recs so please start sharing them!

That makes sense, thankyou!

Do professors notice messages from students on these platforms?

Tbh I am a 2nd year undergraduate rn, but im not using Twitter/bluesky with the sole purpose of getting an internship, but yeah the audience is there.

PS: I've been cold emailing professors for RL research internships, any tips?