VisualScratchpad: Inference-time Visual Concepts Analysis in Vision Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Lim, Hyesu, Choi, Jinho, Kim, Taekyung, Heo, Byeongho, Choo, Jaegul, Han, Dongyoon
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2603.07335