Publication
Inseq: An Interpretability Toolkit for Sequence Generation Models
Gabriele Sarti; Nils Feldhus; Ludwig Sickert; Oskar van der Wal; Malvina Nissim; Arianna Bisazza
In: Danushka Bollegala; Ruihong Huang; Alan Ritter (Hrsg.). Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics: System Demonstrations. Annual Meeting of the Association for Computational Linguistics (ACL-2023), July 9-14, Toronto, Canada, Association for Computational Linguistics, 7/2023.
Abstract
Past work in natural language processing interpretability focused mainly on popular classification tasks while largely overlooking generation settings, partly due to a lack of dedicated tools. In this work, we introduce Inseq, a Python library to democratize access to interpretability analyses of sequence generation models. Inseq enables intuitive and optimized extraction of models' internal information and feature importance scores for popular decoder-only and encoder-decoder Transformers architectures. We showcase its potential by adopting it to highlight gender biases in machine translation models and locate factual knowledge inside GPT-2. Thanks to its extensible interface supporting cutting-edge techniques such as contrastive feature attribution, Inseq can drive future advances in explainable natural language generation, centralizing good practices and enabling fair and reproducible model evaluations.