Welcome AI Research Readers (AIRR)
This is the Github repository for the AI Research readers, focusing on LLM based Agents. We aim to foster a collaborative and closely-knit learning environment through weekly discussions on cutting-edge research on agents.
This is the first time for most of us doing research paper reading group, so please be patient and open to beginners mistakes we will make.
The goal of our reading group is to:
- Take our understanding of AI Agents from 0 to 100.
- Keep abreast of the latest research developments.
- Become a place to learn together and have accountability.
This is an async-first reading group. Here's our approach to ensure everyone gets the most out of our discussions:
- Read the Paper: Start by reading the paper thoroughly. If you're unsure how to approach it, here's a useful guide on how to read a research paper.
- Discussion Preparation:
- Post any questions or points for discussion on the GitHub Discussions tab for that paper at least 2 days before the scheduled call.
- If you're presenting, prepare a document (e.g., PPT, review doc, highlights PDF) to summarize your views on the paper.
- Review all posted questions before the call, and be ready to integrate them into the discussion.
- During the Call: Collaboratively, we'll write a summary document that encapsulates the discussion, key insights, and differing viewpoints.
- Sharing: Feel free to share the insights gained from our discussion outside the group. Promoting shared learning is one of our core missions.
Weekly Discussions:
- Paper Selection and Reading:
- We'll discuss a new paper each week. You can find them in the
paperssection with the schedule below. - We recommend reading the paper by 2 days before our Zoom meeting.
- We'll discuss a new paper each week. You can find them in the
- Discussions on Github:
- We leverage Github Discussions for asynchronous conversations. Find the dedicated thread for the current week's paper in the Discussions section. Feel free to post comments, questions, and insights throughout the week.
- This helps us come prepared for the Zoom meeting, focusing on deeper discussions and clarifying doubts.
- Zoom Meetings:
- We hold weekly Zoom meetings on Fri/Sat at [Meeting Time] (details also in
logistics.md). We'll delve into the paper, discuss key points raised in the Github Discussions, and explore further implications.
- We hold weekly Zoom meetings on Fri/Sat at [Meeting Time] (details also in
Repo Structure:
- This repository is organized to keep discussions focused:\
notes: Each week's paper has a dedicated folder here. These folders may contain:summary.md: A collaborative document summarizing key takeaways from the paper (everyone can contribute). Examplecode: Any code or experimentation notebooks created by readers.
A collection of concise write-ups on each paper, with something noteworthy is also maintained here.
| Date | Topic | Meeting | Presenters | Notes |
|---|---|---|---|---|
| 13th April, 2024 | # The Rise and Potential of Large Language Model Based Agents: A Survey | TBD | Nehil |