Workshop Paper

From Snow to Rain: Evaluating Robustness, Calibration, and Complexity of Model-Based Robust Training

Robustness to natural corruptions remains a critical challenge for reliable deep learning, particularly in safety-sensitive domains. We study a family of model-based training …

Josué Martínez-Martínez
•

Choosing ‘Right’ from Wrong: A Closer Look at Selection Bias in Spatial Multiple-Choice Questions in Large Multimodal Models

This paper examines the limitations of a probing method for determining whether Large Multimodal Models (LMMs) actually understand spatial concepts in image inputs, which may be …

avatar
Giselle Zeno
•

Investigating the Impact of Graph Structure and Attribute Correlation on Collective Classification Performance

Relational machine learning methods can significantly improve the predictive accuracy of models for a range of network domains, from social networks to physical and biological …

avatar
Giselle Zeno
•