IRONIC: Coherence-Aware Reasoning Chains for Multi-Modal Sarcasm Detection

TL;DR

IRONIC introduces a coherence-aware reasoning framework that enhances zero-shot multi-modal sarcasm detection by leveraging cognitive-inspired coherence relations, outperforming existing methods.

Contribution

The paper proposes IRONIC, a novel in-context learning approach that incorporates multi-modal coherence relations for improved sarcasm detection without task-specific fine-tuning.

Findings

Achieves state-of-the-art zero-shot performance on multi-modal sarcasm detection.

Effectively leverages referential, analogical, and pragmatic image-text linkages.

Demonstrates the importance of cognitive insights in multi-modal reasoning.

Abstract

Interpreting figurative language such as sarcasm across multi-modal inputs presents unique challenges, often requiring task-specific fine-tuning and extensive reasoning steps. However, current Chain-of-Thought approaches do not efficiently leverage the same cognitive processes that enable humans to identify sarcasm. We present IRONIC, an in-context learning framework that leverages Multi-modal Coherence Relations to analyze referential, analogical and pragmatic image-text linkages. Our experiments show that IRONIC achieves state-of-the-art performance on zero-shot Multi-modal Sarcasm Detection across different baselines. This demonstrates the need for incorporating linguistic and cognitive insights into the design of multi-modal reasoning strategies. Our code is available at: https://github.com/aashish2000/IRONIC