This project provides a local paper-relation viewer to manage and present relations between papers.
Due to the tremendous growth of research, it is hard for us to keep track of every paper's information. With a graph-based viewer, it will be easier for us to derive the connection between each paper of interest.
-
Graph View
- Present paper as node
- Present edge as similarity
- Zoom in/out
- Show HashTag connections between papers.
- If two nodes' don't have similarity from LLM, the viewer will show the similarity from HashTag.
-
Paper information
- Add paper
- Delete paper
- Edit paper
- Upload PDF
-
Similarity between Papers
- Using hashtag (cosine similarity)
- Using LLM
- Simple usage. (Concate the selected information in on string.)
- Weighted information.
- Include PDFs.
- Customized function.
- Self-defined models.
-
Provide explanations for the edges
- LLM caption
-
Search
- Paper Title
- Venue
- Author
- Hashtag
-
Backup
- Python 3.9–3.13 — PyTorch does not yet support Python 3.14+. If your system Python is 3.14 or newer, install a compatible version first.
“LitAtlas.app” is damaged and can’t be opened. You should eject the disk image.
The downloaded unauthorized application will be quarantined by default.
# replace /Applications/YourAppName.app with actual APP path
# (default will be /Applications/LitAtlas.app)
xattr -cr /Applications/YourAppName.app
