Phillip Isola

Words That Make Language Models Perceive

Better Together: Leveraging Unpaired Multimodal Data for Stronger Unimodal Models

Training Transformers with Enforced Lipschitz Bounds

Fairness Aware Reward Optimization

Single-pass Adaptive Image Tokenization for Minimum Program Search

Personalized Representation from Personalized Generation

Adaptive Length Image Tokenization via Recurrent Allocation

Separating Knowledge and Perception with Procedural Data

On the Cycle Consistency of Image-Text Mappings