Methods to improve my skills in studying papers

[ghaleon]

I'm a PhD student in high-energy theory and this is one point I have been struggling with for a while. When I study a paper what is the most efficient way to do so? For example, should I try to rederive everything following the author's reasoning? Should I try to take notes, and if so how to identify what I should write down? Or should I just read the paper like reading a newspaper? In summary, what methods should I use to be efficient in reading papers on hep-th?

The thing is this: both my master's advisor and my PhD supervisor never gave any advice on the subject and honestly since day one in graduate school it felt like everyone assumed this to be a skill all students already have. I felt ashamed of asking about it (to be honest I'm even ashamed of asking it here, even though I'm not identifying myself), because I feel like this is such a basic skill that it is somewhat disgraceful to ask advice on it.

Still, I'm asking it here now because I really feel I haven't found by myself one efficient method and I must develop this skill. When I try to rederive everything together with the author it seems like it will take forever to read a single paper (and there are quite a few I need to get familiar with, quite a lot of pressure to publish something quickly and so on). The same when I try to take notes, because I really don't identify what I should be writing down and in the end of the day it feels like I'm rewriting the paper. Just reading like a newspaper feels like I'm not really learning the subject, so I really don't know what is the right approach here.

 

[ghaleon]

I'm posting here links with a few papers like the ones I'm asking about (in particular in the subject of flat space holography), just to give an idea of the kind of paper I'm dealing with:

• https://arxiv.org/abs/2105.03516
• https://arxiv.org/abs/2105.09792
• https://arxiv.org/abs/2007.03785
• https://arxiv.org/abs/2005.10666
• https://arxiv.org/abs/2103.03961

 

[paralleltransport]

Not a graduate student anymore, but one thing I found helpful when diving into a new field is to try figure out what are the main seminal paper in the field, study it carefully. If it's relatively new field like the papers you highlight I usually go for the most cited work, or things that are more foundational. From there I would look at references and read only the main conclusions (skip all the proofs) to get an overview of the major results. I'd want a 3-4 sentence summary of each side paper I read (latex them maybe to stay focused).

Over a few weeks I wouldn't know the intricate details of the field, but I would know what are some of the well known results, and have gone through 10+ papers a week. Of course the hard part is narrowing down the list of papers but that part it seems you already figured out.

In a way most papers, I just need to know that:
- they exist and the key problem they solved so I can look it up later.