Keep track of your readings with PaperMemory! It automatically and locally stores papers you open in your browser
When opening a paper (its web page or pdf), a reference to that paper will be stored in your Memory. Supports Arxiv, Open Review, NeurIPS, CVPR, I/ECCV, PMLR, PNAS, ACL, SciRate and more! Ask for new venues on Github.
Click on the Memory button below (or press a )!
⚠️ Important Notice Click here
⋗ Feedback
It's the little things that make your research smoother.
Look at the code ; raise issues ; share improvement ideas
Open your full-page memory to explore papers more easily, in a dedicated tab.
Open the options page to access advanced customization features:
Auto tagging papers from regex
Customize title function
Import papers from lists of URLs
Manage your data (export/load)
Ignore some paper sources
And more!
Syncing
Synced ✅
Sync failed ❌
×
Are you sure you want to delete:?
×
Keyboard Shortcuts
The following shortcuts are available on your browser if and only if they are not already assigned.
Go to your settings or chrome://extensions/shortcuts (Chrome & Brave) or about:addons ➜ Tools for all add-ons cogwheel ➜ Manage Extension Shortcuts (Firefox) to set/edit these shortcuts.
Open the popup
cmd + shift + e on Mac
ctrl + shift + m on Linux & Windows
Trigger manual paper parsing
cmd/ctrl + shift + p
Download current paper's pdf
cmd/ctrl + shift + s
Keyboard navigation
Open the Memory
a from the popup's home will open the Memory
navigate to the bottom left button with tab and hit enter
Search
Search field is automatically focused on memory open
Navigate to the top input with tab or shift + tab
Navigate papers
tab will iterate through papers down the list
shift + tab will go back up the list
Edit a paper
Press e to edit the paper's metadata: tags, code and note when the paper is focused (from click or keyboard tab navigation)
Navigate through fields with (shift+) tab
Press enter to open a focused paper (focus an existing tab with the paper or create a new tab to the paper's pdf if it's not open already)
backspace to delete a focused paper (a confirmation will be prompted first don't worry 👮♀️)
Close Memory or Menu
esc closes the memory (or the menu -- not in Firefox)
Search
In a paper's authors, title and note.
Split queries on spaces: gan im will look for: all papers whose (title OR author) contain ("gan" AND "im")
In a paper's code link
Start the search query with c: to only search code links
In a paper's tags
Start the search query with t: to filter by tags
t: gan will look for all papers whose tag-list contains at least 1 tag containing "gan"
t: gan tim will look for all papers whose tag-list contains (at least 1 tag containing "gan") AND (at least 1 tag containing "tim")
Search paper by years with y: ${year}
Use , to separate requested years or start with > or < to filter paper published after/before a given year (stricly)
y: 20,21,22 will display papers published in 2020 OR 2021 OR 2022
y: <2015 will display papers published before (strictly) 2015
y:>19 will display papers published after (strictly) 2019
[feature] You can now parse arbitrary websites: you just need to confirm / update the information PaperMemory was able to extract from the webpage
[new source] Papers from huggingface.co/papers/ will now be added as arxiv papers. If PaperMemory discovers a publication venue, a badge will be displayed on the HF paper page
[new source] Thanks to @julien-blanchon you can also automatically go to ar5iv or arxiv-vanity paper pages. Those will be stored as Arxiv papers.
[dev] Various bug fixes in the notifications and parsing procedures
[feature] You can now include the venue and year of a paper in its Markdown link (see popup menu)
[feature] Deduplication: PaperMemory now matches and merges papers based on their titles not to create multiple Memory entries for the same paper
[feature] Paper imports: in the advanced options page, in the Data Management section, you can now import papers from a list of urls
[feature] Paper exports by tag: in the same Data Management section, you can now select papers to export by providing a list of tags and obtain a .bib or .json file with the appropriate papers.
If you export papers by tags and give the file to someone on your team, they'll be able to import them!
[feature] In the menu, you can now disable automatic paper recording. In which case you can trigger paper parsing manually from the popup or with a keyboard shortcut ctrl/cmd + shift + p
[feature] There's a new keyboard shortcut to download a paper (it goes to your PaperMemoryStore if you have enabled the feature) cmd/ctrl + shift + s
[dev] More tests
[dev] Fix paper parsing from pdf urls for Nature and IJCAI venues
[feature] In an effort to use as few as possible arxiv.org references, PaperMemory will use dblp.org and crossref.org bibtex data whenever possible for your published papers' citations. In other words it tries to discover published versions of you pre-prints and this will be effective when copying/exporting bibtex data
[feature] New configuration option to only store official repositories when discovering code with PapersWithCode
[feature] New configuration option to prefer using links to abstract pages vs pdfs
[feature] New configuration option to only record paper from pdfs, not their online pages/abstracts
[feature] New full-page memory (link at the bottom of the menu) to explore your memory in a fill-size and dedicated tab.
[feature] New advanced options to filter out default paper sources.
[feature] You can now also sort your memory by paper upload/publication year
[feature] You can now search papers by year (y: 18,19), including before/after a date (y: <2019, y: >2018).
[feature] Weekly local backups (up to 6 weeks) in case of a bug/user mistake (see Readme on Github)
[new source] Add Nature papers parsing
[new source] Add ACS papers parsing
[new source] Add IOPscience papers parsing
[new source] Add JMLR papers parsing
[dev] New permission to monitor all urls. This is necessary because otherwise every new paper source will be associated with a new permission, disabling the extension until you re-enable it manually.
[dev] Performance improvements, notably a step towards losing the JQuery dependency
[dev] Improve & standardize Bibtex parsing, formatting and exporting
[dev] Fix PNAS paper parsing (due to website update)
[dev] Improve search to include tags (still t:tagName to search tags only)
Features Arxiv enhancements: titles, bibtex & markdown links
Manual Parsing
Automatic, url-based parsing has been disabled.
To manually add papers you must now setup a keyboard shortcut to trigger this action.
Visit your browser's settings to set it.
This paper is not in your memory, and it can be for one of many reasons:
You disabled paper recording from non-pdf pages in the menu
You disabled this paper source in the options page
You deleted the paper (refresh the page to add it back)
Your memory was synced and the remote data did not have the paper (it was deleted on another device) (refresh the page to add it back)
There was an error parsing the paper's data (you can check the console if you think this is an issue)
On Firefox, content scripts are not triggered on pdfs.
This is not something I can do anything about, it's a design choice by Firefox developers.
The extension would work on the paper's abstract
You are actually not on a paper page but the extension made a mistake thinking so, just ignore this.
Open an issue on Github if you think you encountered a malfunction.
Deprecation notice: custom title function
PaperMemory displayed a warning notification because it has detected that you are using a custom title function.
If it is not the case, meaning you have not changed the function that produces a title from a paper in your Advanced Options, you can safely ignore this warning.
If you do however use a custom title function, be aware that this functionnality will be removed in version 0.6.0.
Why?
Mostly because of safety, code quality and accessibility.
The vision for this feature is to make it modular but not code-based. Something like asking you for a list of attributes you want to be part of the title and potenlially lower/upper case or capitalizing.
This vision is not fully-shaped but I wanted to give you a proper heads-up to give you time to adjust.
What should you do about it?
If you're happy with the vision for this feature, you don't have anything to do about it.
If you want to keep the feature as is, let's go and discuss on Issue #109!