Attribution Bias in LLMs
Published: 7 July 2025| Version 1 | DOI: 10.17632/g65cxm6s99.1
Contributor:
蕾蕾 蒋Description
A controlled experiment involving two event types, two attribution perspectives, four social attributes, and prompt-based interventions was conducted across five LLMs, including ChatGPT and four leading Chinese models.
Files
Categories
Explainable LLM