Zachary C. Lipton

Analyzing how red-teaming can enhance AI safety and address potential risks.

This article discusses the impact of data under-reporting on fairness in predictive models.

New datasets reveal difficulties in identifying racers and text in muddy conditions.

A new method enhances person recognition using weakly labeled data.

This article explains the phenomenon of hallucinations in image generation models.

Discover how machines learn without forgetting using synthetic data and expert systems.