DBDM: a formal framework for bias detection and mitigation in distributed large language models
Large language models (LLMs) are increasingly being deployed in distributed settings, where different nodes may work on heterogeneous data, user populations, and local model updates. In such settings, bias can arise in an emergent fashion via local interactions and spread over the connected nodes, turning fairness into...
K. J. Sahana Devi, Nishant Tripathi, S. Santhosh Kumar et al.
· Scientific Reports · 0 citations