Skip to ContentSkip to Navigation
About us Practical matters How to find us G. (Gabriele) Sarti, MSc

Research interests

I am a PhD student at the Computational Linguistics Group of the University of Groningen and member of the InDeep consortium, working on user-centric interpretability for neural machine translation. I am also the main developer of the Inseq library. My supervisors are Arianna Bisazza, Malvina Nissim and Grzegorz ChrupaƂa. My research aims to empower users of neural machine translation systems by developing interpretable and interactive tools for more efficient professional post-editing.

My research interests focus on interpretability for NLP models, in particular to the benefit of end-users and by leveraging human behavioral signals. I am also passionate about language games, uncertainty estimation and open source collaboration.

Publications

A Primer on the Inner Workings of Transformer-based Language Models

Are Character-level Translations Worth the Wait? Comparing ByT5 and mT5 for Machine Translation

DecoderLens: Layerwise Interpretation of Encoder-Decoder Transformers

Democratizing Advanced Attribution Analyses of Generative Language Models with the Inseq Toolkit

IT5: Text-to-text Pretraining for Italian Language Understanding and Generation

Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation

Multi-property Steering of Large Language Models with Dynamic Activation Composition

Non Verbis, Sed Rebus: Large Language Models Are Weak Solvers of Italian Rebuses

Contrastive Language-Image Pre-training for the Italian Language

DecoderLens: Layerwise Interpretation of Encoder-Decoder Transformers

Press/media

Can Word-level Quality Estimation Inform and Improve Machine Translation Post-editing?

The AI that explains images in Italian

CLIP-Italian: A new AI model to connect images and texts in Italian