My favorite papers are those that clearly explain their thought process, show what they did to a great level of detail, and in a clean and easy to read format.

My favorite papers are those that clearly explain their thought process, show what they did to a great level of detail, and in a clean and easy to read format. These papers can be used as a sort of make shift tutorial, which can be incredibly helpful when studying an unfamiliar topic. (I actually like them better than tutorials, because I get some perspective into their thinking and how they came to explore the topic which makes it easier to generalize the lesson.) Not just reading papers, but Wikipedia as well: When exploring a new domain of knowledge, you may not know much or any of the vocabulary being used. This turns the average person off. I see this as the most common hurdle faced by those self studying a new topic. The trick is to slow down, look at the different vocabulary you don't recognize, and then look them up. What sometimes happens is to learn this new vocabulary you'll be met with a wall of even more unfamiliar vocabulary. This turns into a dependency chain of unfamiliar topics. Looking up those words / dependencies recursively, eventually you'll get to familiar knowledge. You can start building up from that place of familiarity and before you know it you've picked up not only the basics of a new domain of knowledge but the more advanced topics build up to what you were interested in. Learning the prerequisites usually takes hours, but sometimes this process can take days. Taking days to learn a dependency chain to learn an advanced topic seems like a long time, but in comparison to taking a class on the topic or reading a text book, this style of learning is highly accelerated. Sometimes there is no alternative learning style if you need to parse a paper on a cutting edge unfamiliar topic not yet covered in text books, and imo if you pace yourself and value learning over any other goal, it can be quite a relaxing and enjoyable process. Reading papers is a lot of fun. When I'm learning new topics, I take notes. My goal is to maintain what I learn, not have it go in one ear and out the other, so why not slow down, relax, and take notes? Better yet if I come back to the same topic months later I can look over my notes. If I forgot something about the topic, I know my study style could be improved. This meta-learning feedback loop can be super helpful if you value and enjoy learning new topics (which you do or you wouldn't be on YC).