📜 Notion template now available here.
Authors: [Authors]
Date Published : [Date published or submitted.]
Paper Link : [Personal annotated version of the paper or public online version. ]
Talk Link: [Optional talk link e.g. from slideslive if published.]
Comments: [e.g. Published at X / arXiv paper/ in review.]
TLDR: [One or at most two line summary.]
Relevance: [Score between 1 and 5, stating how relevant this paper is to your work. Usually filled in at the end. ]
Tags: [General or specific research topic tags e.g. rl, nlp or efficient-nlp. Could also be used to specify the conference this was accepted at or any useful grouping.]
Soundness: [Score between 1 and 3 for how sound the results in the paper are. Was their training and/or evaluation settings fair and sound? Did they fairly compare to all relevant work? Similiar to the soundness scores for conference reviews.]
Main Takeaway(s): [Main takeaway(s) from the paper. Usually 1 or two sentences max. This usually only makes sense if you have done a detailed read of the paper. I would be mindful of assumptions/soundness of the paper when deciding what the main takeaways are. ]
[Summary of the paper - a few sentences with bullet points. What did they do? What is their hypothesis? ]
[What are the issues that the paper addresses? Describe the problem. Why did they write this paper?]
[ Only for work that appears to be highly related from reading abstract/intro/conclusion.]
[ What is the problem setting e.g. regression, classification or sequence prediction? In RL, this would also include the environment and details about rewards etc. This should include the evaluation setting for the results. ]
[How did they approach the problem. What methods did they use?]
[What assumptions were made and are these assumptions valid?]
[Can be empty]
[Theoretical or empirical results (any main graphs and tables). Also try to possibly mention why you or the authors think certain results occurred. ]
[Did the authors mention any reservations/limitations to their work or methodology? Do you see any limitations of their work?]
[Is there anything that is confusing and could need better explanations or references?]
[What is the author's conclusion? What do they claim about their results?]
[What do you think about the work presented in the article? Did the authors manage to achieve what they set out to achieve?]
[Fine, Good, Great]
[Can you think of ways to improve this paper or ideas for future work?]
[If related to own work.]
- What can we learn from their approach:
- How are we different:
- Cited references to follow up on/related papers/check google scholar for papers citing this paper:
- Source code/blog/twitter thread/other links: