The "Literature Review" part of academic papers can take a while to write, often misses some related sources, and doesn't keep up with new publications. LitGPT is a proof of concept showing how AI can change the way we do literature reviews. Simply provide your paper's title and abstract to see the results!
- 📚 Right now LitGPT only covers ML and AI papers submitted on arxiv from 2012 to 2022 🗓️
- 🚀 The output quality is much better with GPT-4 if your API_Key has access to it 🔑
- 📄 LitGPT only uses titles and abstracts. The output quality is significantly better with the whole paper text, but the data should be extracted from pdf files and take much more space and memory 💾
- 🔢 The number of papers is limited to 15 for ChatGPT and 25 for GPT-4 📊
2p7sByqvlQXii4BB.mp4
- To get the paper data (might take a while, I can't upload the data because of github size limits)
python api/get_data.py- To run the app
export FLASK_APP=api/index.py
flask run