H^2em: Enhancing Zero-Shot Learning with Hierarchical Hyperbolic Embeddings

This research explores the use of hierarchical hyperbolic embeddings to improve compositional zero-shot learning, a critical area in AI. The study's focus on zero-shot learning suggests a potential advancement in models' ability to understand and generalize to novel concepts.

• •Focuses on zero-shot learning, allowing models to generalize to unseen data.
• •Utilizes hierarchical hyperbolic embeddings for representation.
• •Potentially improves compositional understanding of new concepts.