SynSciPass: A dataset of synthetic scientific texts for detecting appropriate and inapproriate uses of text generation
This repo accompanies the paper (SynSciPass: detecting appropriate uses of scientific text generation)[] and contains the SynSciPass dataset (./data)
The SynSciPass dataset is contained in ./data with the train, test, and validation splits at synscipass_.
Requires python 3.8 or higher.
$ pip install -r requirements.txt
Note: all identifying information is overwritten with anon.
Run the notebooks in the SynSciPass creation folder.
Run the notebooks in the DAPT-TAPT and Train SynSciPass folders.
Run the notebooks in the Experiment # folders.