LLM Benchmarking

data.zip: All the questions that used in the expirements saved for reproducibility reasons rephrased_data.zip: All the rephrased questions that used in the expirements saved for reproducibility reasons results.zip: Contains all LLMs answers on exact/original questions rephrased_results.zip: Contains all LLMs answers on rephrased questions csv.zip: Tables for descriptive statistics score (mean, stdev) for 4 datasets on all models output_plots.zip: Performance shifts plots beta_binomial.zip: Bayesian results for kappa, mu, gamma, svg images codebase.zip: All the source code for download, preprocess data, call the LLMs, create the results, run the Bayesian framework and create the plots. All are described in detail inside the README files.