├── dataset : direcotry for experiment datasets
├── README.md : introduction
├── calculate_sim.py : code for calculating kendall's tau
├── cut_figure.py : code for cutting figures for writing
├── draw.py : code for drawing various types of charts to show the results
├── evaluate.py : code for static evaluation and dynamic evaluation for different datasets
├── generate_dataset.py : code for dataset generation/sampling
├── methods.py : code for some general functions
├── moderation.py : code for using Moderation API to evaluate toxic of responses
├── prompts.py : system prompts
├── run.py : code for runing with one-shot
├── run_adv.py : code for runing experiments on advbench dataset
├── run_bias.py : code for runing experiments on bias dataset
├── run_csqa.py : code for runing experiments on csqa dataset
├── run_fact.py : code for runing experiments on fact dataset
└── run_gsm8k.py : code for runing experiments on gsm8k dataset
- Replace the OPENAI api key with your own key in in function "get_client" in methods.py
- Set parameters in run.py
- Run run.py