TRACE: A Framework for Analyzing and Enhancing Stepwise Reasoning in Vision-Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Imani, Shima, Moon, Seungwhan, Mathias, Lambert, Zhang, Lu, Damavandi, Babak
Format:	Preprint
Published:	2025
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/2512.05943