TY  - RPRT
TI  - Inverse-LLaVA: Rethinking Multimodal Alignment via Text-to-Vision Mapping
AU  - Xuhui Zhan
AU  - Tyler Derr
PY  - 2026
UR  - https://arxiv.org/abs/2508.12466
ID  - 2508.12466
ER  -