We study how in-context learning (ICL) in language models is affected by semantic priors
versus input-label mappings. We investigate two setups-ICL with flipped labels and ICL with …

Abstract Machine learning has advanced dramatically, narrowing the accuracy gap to
humans in multimodal tasks like visual question answering (VQA). However, while humans …

Abstract We introduce Three Towers (3T), a flexible method to improve the contrastive
learning of vision-language models by incorporating pretrained image classifiers. While …

We study the impact of different loss functions on lesion segmentation from medical images.
Although the Cross-Entropy (CE) loss is the most popular option when dealing with natural …

Recent work has shown the importance of reliability, where model performance is assessed
under stress conditions pervasive in real-world deployment. In this work, we examine …

This work examines the claim in recent work that Bayesian neural networks (BNNs) are
inherently robust to adversarial perturbations. To study this question, we investigate whether …

We study how in-context learning (ICL) in language models is affected by semantic priors
versus input–label mappings. We investigate two setups—ICL with flipped labels and ICL …

Data has powered incredible advances in machine learning (ML). Yet, the kinds of data
used for training are often hard labels aggregated over humans' annotations, which fail to …