ELI-Why

Benchmark and Code for ELI-Why

The ELI-Why benchmark is stored as ELI_Why.jsonl.

Project Structure

• RQ1/
• Individual Metrics/ Contains scripts for running individual metrics and analyses. To run individual metrics, navigate to the RQ1/ directory and execute:

python3 individual_metric_readability_analysis.py --model GPT4o --set full

Alternatively, to run all combinations, use:

python3 individual_metric_readability_analysis.py --all

• TESDIFF(Similarity) Metric/ Contains scripts for running TESDIFF metric and analyses. To run it, in the RQ1/ directory and execute:

python3 tesdiff.py --all