Revelio: Interpretable Long-Form Question Answering

Gianluca Moro, Luca Ragazzi, Lorenzo Valgimigli, Fabian Vincenzi, Davide Freddi

Keywords: Long-Form Question Answering, Pretrained Language Models, Interpretability, Knowledge Graph, Natural Language Processing

TL;DR: We present Revelio, a new plugin layer that enables PLM to interact with a KG to provide interpretable reasoning paths as answer rationales in long-form question answering.

Abstract: The black-box architecture of pretrained language models (PLMs) hinders the interpretability of lengthy responses in long-form question answering (LFQA). Prior studies use knowledge graphs (KGs) to enhance output transparency, but mostly focus on non-generative or short-form QA. We present Revelio, a new layer that maps PLM's inner working onto a KG walk. Tests on two LFQA datasets show that Revelio supports PLM-generated answers with reasoning paths presented as rationales while retaining performance and time akin to their vanilla counterparts.

Submission Number: 142