TY  - RPRT
TI  - FOCUS: Internal MLLM Representations for Efficient Fine-Grained Visual Question Answering
AU  - Liangyu Zhong
AU  - Fabio Rosenthal
AU  - Joachim Sicking
AU  - Fabian Hüger
AU  - Thorsten Bagdonat
AU  - Hanno Gottschalk
AU  - Leo Schwinn
PY  - 2025
UR  - https://arxiv.org/abs/2506.21710
ID  - 2506.21710
ER  -