一作ACL 2026Findings· Findings of ACL
Two-Stage Parameter Alignment for Multi- L o RA Merging in Large Language Models
Zijian Li · Xiachong Feng · Weitao Ma · yichong huang · Xiaocheng Feng · Bing Qin
摘要
Merging a large number of low-rank adaptations (LoRAs) is a key technology for enhancing the integration and deployment efficiency of large language models (LLMs). However, current general model merging methods are prone to “parameter interference” problem, and this issue is especially pronounced...