Mark Vero

A new benchmark for evaluating fairness in representation learning methods.

Examining the dangers of quantized language models and their potential misuse.

Analyzing vulnerabilities in popular code completion tools and their implications for developers.