ProCrit: Self-Elicited Multi-Perspective Reasoning with Critic-Guided Revision for Multimodal Sarcasm Detection

2026-05-20

AI Summary

This work addresses the challenge of multimodal sarcasm detection, which hinges on modeling cross-modal incongruity between literal expressions and intended meanings. Existing approaches are limited by fixed, predefined perspectives that fail to capture the diverse mechanisms underlying sarcasm. To overcome this, the authors propose ProCrit, a novel dual-agent collaborative framework in which a proposer agent dynamically generates sample-specific analytical perspectives, and a critic agent provides natural language feedback to drive an iterative draft-critique-revise reasoning process. Integrating dynamic role generation, autoregressive sequence modeling, and bidirectional reinforcement learning, ProCrit enables self-generated perspectives and feedback-driven co-optimization. Extensive experiments on three mainstream multimodal sarcasm detection benchmarks demonstrate significant performance gains, confirming the framework’s effectiveness and generalization capability.

Abstract

Multimodal sarcasm detection requires reasoning over cross-modal incongruities between literal expression and intended meaning, yet the specific analytical perspectives needed vary across samples due to the diversity of sarcastic mechanisms. While recent methods make this analytical process explicit, they still rely on fixed, predefined perspectives that operate independently under hand-crafted routing rules. We argue that multimodal sarcasm detection instead calls for self-elicited multi-perspective reasoning, where a model autonomously generates the perspectives needed for each sample and progressively integrates them into a coherent analysis. To realize this goal, we propose ProCrit, a Proposal-Critic two-agent framework with a proposal agent for multi-perspective reasoning and a critic agent for external evaluation and targeted revision guidance. First, to overcome the lack of process-level supervision in existing sarcasm datasets, ProCrit synthesizes process-level reasoning annotations through a dynamic-role agentic rollout: a strong vision-language model sequentially spawns analytical roles within a shared context, and the resulting multi-role trajectories are flattened into sequences that preserve cross-perspective dependencies while enabling efficient autoregressive generation. Second, to improve reasoning reliability, ProCrit adopts a draft-critique-revise paradigm in which an independent critic identifies reasoning deficiencies and provides targeted natural-language feedback for directed revision. Finally, we develop a mutual-refinement training framework that jointly optimizes proposal drafting and feedback-guided revision via dual-stage reinforcement learning, while refining the critic agent according to the actual effectiveness of its feedback. Experiments on three widely used benchmarks demonstrate the effectiveness of ProCrit.

Problem

• multimodal sarcasm detection
• cross-modal incongruity
• multi-perspective reasoning
• analytical perspectives
• sarcasm mechanisms

Innovation

• self-elicited reasoning
• multi-perspective analysis
• critic-guided revision
• process-level supervision
• agentic rollout