From Insights to Impact: Actionable Interpretability for Neural Machine Translation | Gabriele Sarti

From Insights to Impact: Actionable Interpretability for Neural Machine Translation

Gabriele Sarti, Willem Zuidema, Ivan Titov, Barbara Plank, Eva Vanmassenhove
Natural Language Processing, Academic
Project Project Slides

Date
Dec 11, 2025
Event
Workshop on Actionable Interpretability for Language Models and Machine Translation Systems
Location
House of Connection, University of Groningen
Groningen, Netherlands
Natural Language Processing Interpretability Language Modeling Feature Attribution Steering Sparse Autoencoders Machine Translation Post-editing User study

Related