Arianna Bisazza

Study reveals bias differences in language models across various languages.

Mirage enhances answer attribution in retrieval-augmented generation systems.

A new benchmark assesses how well language models follow multiple instructions in sequence.

Research on group interactions in language learning using advanced neural agents.

This study analyzes how well AI models solve Italian rebuses.

Learn how the sequence of information affects AI answer quality.

Research shows variation in speech improves language model training.

Language models need training to handle biases and toxicity in different languages.