Sara Hooker

A new method enhances translation quality by optimizing training data selection.

Introducing IrokoBench to improve LLM evaluation in African languages.

This article discusses methods for improving AI alignment with various cultures.

Exploring how synthetic data shapes machine learning models and their behavior.

New methods for training language models to support diverse languages.

Studying how quantization affects performance in different languages.

Explore the need for an open feedback system to improve AI responses.

This article discusses a new method for generating synthetic data using various models.

Nexus combines efficiency, specialization, and adaptability in language model development.

Exploring concerns of AI impacting biological safety and management.

Addressing cultural biases in multilingual evaluation for better language model performance.