Topics

perplexity.ai

elicit.org : more for research

consensus: evidence-based answers

Papers and Demos

Steps to read the papers

  1. Search for the papers (sites below)

Some sites for AI/ML papers

  1. Check the credibility of the paper:

<aside> đź’ˇ Tool to figure out if an article is worth reading or not by visualizing the citations:

</aside>

Connected Papers | Find and explore academic papers

  1. First pass: For ALL the papers you have chosen so far, read the title, abstract, keywords, and conclusion. See if they are really about what you are looking for. It will give you a basic idea about the paper and will help you decide whether you want to keep on reading it or not.

  2. Second pass: For the papers that passed the first round, go a bit more in-depth. Look at the graphs and tables, read their captions. You can also quickly go through the introduction and related works to see if you find it interesting and well done, but don’t dive into the method and experiments right away. This takes time to digest and understand. You need to be sure it is the right paper for you. This second pass will help you get the crisp of the paper, and by then, you will already be able to summarize it and the results.

This might be helpful as well:

Explainpaper

<aside> 💡 Before the third pass: Get a background on these papers for efficiency. YouTube has a lot of channels that explain these research papers. E.g., people like Yannic Kilcher, What’s AI, Letitia. Maybe I’ll find more people that are of my taste.

OR

Using Medium is a great way to find paper summaries and great explanations, either on Towards AI or Towards Data Science publications.

OR

To get a rapid overview without diving into the theory, the classic Two Minute Papers does a fantastic job as well.

</aside>