Reproduce “Analyzing Traditional and Neural Approaches to Multilingual Readability Assessment” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Joshua Wong, Chris Tanner
The authors test whether features used by traditional readability classifiers align with transformer predictions of CEFR levels for sentences in five languages. They use SHAP to select features and TCAV to assess directions in model representations. Such patterns appear in at least one model family in four languages; Arabic shows separability without consistent alignment, within this dataset and setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.