Research Papers Hub

Explore our comprehensive collection of research papers covering various topics related to the Value Compass project. Stay informed and delve into the latest academic insights and findings from our team..

    Alignment Dataset

    Value FULCRA - Mapping Large Lnguage Modeuls to the Multidimensional Spectrum of Basic Human Values

    Evaluation Dataset

    DenEvil - Towards Deciphering and Navigating the Ethical Values of Large Language Models via Instruction Learning

    Analysis

    Unpacking the Ethical Value Alignment in Big Models

    Analysis

    From Instructions to Intrinsic Human Values —— A Survey of Alignment Goals for Big Models