Rationalizing Transformer Predictions via End-To-End Differentiable Self-Training

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Brinner, Marc, Zarrieß, Sina
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918125639827456
author Brinner, Marc
Zarrieß, Sina
author_facet Brinner, Marc
Zarrieß, Sina
contents We propose an end-to-end differentiable training paradigm for stable training of a rationalized transformer classifier. Our approach results in a single model that simultaneously classifies a sample and scores input tokens based on their relevance to the classification. To this end, we build on the widely-used three-player-game for training rationalized models, which typically relies on training a rationale selector, a classifier and a complement classifier. We simplify this approach by making a single model fulfill all three roles, leading to a more efficient training paradigm that is not susceptible to the common training instabilities that plague existing approaches. Further, we extend this paradigm to produce class-wise rationales while incorporating recent advances in parameterizing and regularizing the resulting rationales, thus leading to substantially improved and state-of-the-art alignment with human annotations without any explicit supervision.
format Preprint
id arxiv_https___arxiv_org_abs_2508_11393
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Rationalizing Transformer Predictions via End-To-End Differentiable Self-Training
Brinner, Marc
Zarrieß, Sina
Computation and Language
Machine Learning
We propose an end-to-end differentiable training paradigm for stable training of a rationalized transformer classifier. Our approach results in a single model that simultaneously classifies a sample and scores input tokens based on their relevance to the classification. To this end, we build on the widely-used three-player-game for training rationalized models, which typically relies on training a rationale selector, a classifier and a complement classifier. We simplify this approach by making a single model fulfill all three roles, leading to a more efficient training paradigm that is not susceptible to the common training instabilities that plague existing approaches. Further, we extend this paradigm to produce class-wise rationales while incorporating recent advances in parameterizing and regularizing the resulting rationales, thus leading to substantially improved and state-of-the-art alignment with human annotations without any explicit supervision.
title Rationalizing Transformer Predictions via End-To-End Differentiable Self-Training
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2508.11393