Leshem Choshen

A new method to assess large language models using fewer examples.

A new benchmark assesses language models' understanding of linguistic competence.

A look at how AI models grasp essential knowledge of the world.

Research outlines techniques to improve efficiency in serving LoRA adapters.

This paper presents a method to assess language models across various prompts.

A study on collecting and using user feedback to improve language models.

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

Share your AI chats easily and help improve models.

This study analyzes how language models share knowledge in different languages.

Researchers examine the reliability of metrics for language model safety.

ZipNN compresses AI models efficiently, keeping essential details intact.

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

A competition aimed at improving how machines learn languages like children do.

Learn how Sloth is changing predictions for language model performance.