V-LUMEN: Visual Lookup Memory for Embedding Scaling in Vision-Language Modelswith Miso Choi, Eunji Kim, Jungbeom Lee2 versionsW_bskx62d8v2 · current