Heterogeneous Value Alignment Evaluation for Large Language Models

Zhaowei Zhang, Ceyao Zhang, Nian Liu +5

The emergent capabilities of Large Language Models (LLMs) have made it crucial to align their values with those of humans. However, current methodologies typically attempt to assig…