作者位次 #2/6ACL 2026Poster· Natural Language Processing
S har V e T : Similarity-aware Parameter Sharing with Vector-based Tuning for Efficient LLM Compression
Jeongin Yun · Jaeri Lee · Jongjin Kim · Minjun Kim · Jinho Song · U. Kang
摘要
How can we share parameters within large language models to significantly reduce memory costs while preserving accuracy? While parameter sharing is a promising solution to the memory overhead of large language models, existing methods rely on naive grouping and fail to correct sharing-induced dis...