Parameter Alignment Mitigates Catastrophic Forgetting in Multilingual Expert Language Models 事件

Parameter Alignment Mitigates Catastrophic Forgetting in Multilingual Expert Language Models arXiv:2606.00284v1 Announce Type: new Abstract: While continual pretraining~(CPT) is a practical way to extend large language models to new languages, na\"ive finetuning on targeted data erodes existing capabilities through catastrophic forgetting. Organizing training around language families reduces cross-language interference but cannot alone prevent forgetting of the general knowledge needed for down