Interpretable zero-shot stance detection with proactive content intervention

Apoorva Upadhyaya

doi:10.1016/j.ipm.2025.104223

Interpretable zero-shot stance detection with proactive content intervention

2025-06-16

0

OA

AI

DOI：10.1016/j.ipm.2025.104223

Original

Original request for help

Save

Abstract

En 中文

• An interpretable approach with rational decoding and LLMs aids zero-shot stance task. • Intentions with toxic, moral, temporal, and emotional aspects help decode the stance. • Results on benchmark datasets validate the effectiveness of our interpretable design. • LLM interventions better address stereotypes with empathy and persuasive impact.

Keywords:

Zero-shot stance detection

Intervention

Interpretability

Rationale decoding

Large language models

Interpretable zero-shot stance detection with proactive content intervention

Abstract

Journal

Organization