00:00
2026-06-22
aclanthology.org
large-language-models
A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM𝛥 Integration into Upcycled MoE
Researchers introduced PARAMΔ, a method that upcycles a dense LLM into a Mixture-of-Experts (MoE) architecture to expand language capabilities without costly alignment. By grafting a MoE-expanded para…