Make Your Decision Convincing! A Unified Two-Stage Framework: Self-Attribution and Decision-Making

Yanrui Du, Sendong Zhao, Haochun Wang, Yuhan Chen, rui bai, Zewen Qiang, Muzhen Cai, Bing Qin

Submission Type: Regular Long Paper

Submission Track: Question Answering

Submission Track 2: Interpretability, Interactivity, and Analysis of Models for NLP

Keywords: rationale, reliable link, two-stage framework

TL;DR: Our proposed SADM framework establishes a more reliable link between the generated rationale and model decision while improving task performance and rationale quality.

Abstract: Explaining black-box model behavior with natural language has achieved impressive results in various NLP tasks. Recent research has explored the utilization of subsequences from the input text as a rationale, providing users with evidence to support the model decision. Although existing frameworks excel in generating high-quality rationales while achieving high task performance, they neglect to account for the unreliable link between the generated rationale and model decision. In simpler terms, a model may make correct decisions while attributing wrong rationales, or make poor decisions while attributing correct rationales. To mitigate this issue, we propose a unified two-stage framework known as Self-Attribution and Decision-Making (SADM). Through extensive experiments on five reasoning datasets from the ERASER benchmark, we demonstrate that our framework not only establishes a more reliable link between the generated rationale and model decision but also achieves competitive results in task performance and the quality of rationale. Furthermore, we explore the potential of our framework in semi-supervised scenarios.

Submission Number: 5011