Modern-day LLMs are "fiction machines," designed not to be truthful but to make sense. What can we expect from these machines, and what are their limitations?

Abstract: Compositionality, or correctly recognizing scenes as compositions of atomic visual concepts, remains difficult for multimodal large language models (MLLMs). Even state of the art MLLMs such ...

Abstract: Compositionality enables humans to learn new concepts from their components and the way they combine. Spatial relationships describe how objects are combined in the real world. These are ...

Mastering a rich repertoire of motor behaviors, as humans and other animals do, is a surprising and still poorly understood outcome of evolution, development, and learning. Many degrees-of-freedom, ...

Compositional data (CoDa) are prevalent in environmental research. They represent parts of a whole, such as percentages, proportions, and relative or absolute abundances. They are arrays of positive ...

This repository contains my solutions of the assignments of the Stanford CS224N: Natural Language Processing with Deep Learning course from winter 2022/23. There are many other great repositories on ...

CVPR Learning Conditional Attributes for Compositional Zero-Shot Learning paper/code ICLR Learning to Compose Soft Prompts for Compositional Zero-Shot Learning paper/code ICCV Hierarchical Visual ...