works on
in-context learning 1knowledge updating 1large language models 1low-rank adaptation 1memory mechanisms 1retrieval-augmented generation 1

From the 1 of 4 linked papers with an AI index.

collaborators
Showing cs.LGShow all

1 paper · 1 filter

cs.LG2026

Understanding LoRA as Knowledge Memory: An Empirical Analysis

Seungju Back, Dongwoo Lee, Naun Kang +4

The paper empirically studies Low-Rank Adaptation (LoRA) as a parametric memory for updating knowledge in large language models, analyzing its capacity, composability, and scaling…