This is the official code repo for the paper: Heterogeneous Value Alignment Evaluation for Large Language Models
All tasks related prompts used in HVAE are listed in prompts
folder, you can selcect claude_style prompt or gpt_style prompt according to the under evaluation LLM model.
The code under utils
foler are the tools related with SVO measurement.
HVAE is released under MIT License.