Semantic Composition in Visually Grounded Language Models

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.