IRON-Meme：结构化理解网络迷因叙事中身份与角色的全新基准
英文标题：IRON-Meme: A Benchmark for Structured Understanding of Identities and Roles in Online Meme Narratives
针对社交媒体中政治迷因（Meme）对网络叙事的塑造与极化影响，研究团队提出了基于社会科学理论的全新评测基准 IRON-Meme。该基准包含 1,110 张经过人工细粒度标注的政治迷因，系统性覆盖社会身份、内/外群体视角与叙事角色三大维度。基于 9 款视觉语言模型（VLM）的评测显示，前沿规模与推理模型性能最佳，但在低频标签上准确率明显下降，且直接引入模板先验知识未能增强叙事角色的分类能力。该研究为开发者和多模态 AI 领域提供了深入理解网络复杂亚文化与多模态修辞叙事的重要评估工具，相关数据及代码已开源。
2.0 KB
640 B
612 B
380 B
24.2 KB
34.7 KB
中文
[提交于 2026 年 10 月 22 日]
标题：IRON-Meme：结构化理解网络迷因叙事中身份与角色的全新基准
查看 Gabriella Bollici 等 4 位作者撰写的论文《IRON-Meme: A Benchmark for Structured Understanding of Identities and Roles in Online Meme Narratives》PDF 版本
摘要：随着社交媒体极化现象的加剧，理解政治迷因（Meme）如何塑造网络叙事变得日益关键。以往研究要么依赖难以规模化扩展的定性分析框架，要么依赖缺乏理论基础的计算方法。这两种途径都孤立地对待叙事类别，且仅聚焦于身份的道德化描绘。为弥补这些局限，我们提出了一个植根于社会科学理论的全新分析框架，涵盖三个核心类别：社会身份（如性别、政治立场与族裔）、视角（内群体与外群体视角）以及叙事角色（道德品格、智商、地位与权力）。在此框架之上，我们推出了 IRON-Meme 基准——这是一个包含 1,110 张经人工标注的政治迷因数据集，涵盖的模板数量显著多于现有基准。我们对 IRON-Meme 的分析表明，绝大多数迷因带有贬损性角色塑造，且身份、角色与模板之间存在系统的关联性。对 9 款视觉语言模型（VLM）的评测表明，前沿规模模型与推理模型取得了最佳性能，但在低频标签上的社会身份与叙事角色准确率有所下降。值得注意的是，引入模板背景知识并未改善叙事角色的分类效果，这表明对迷因的解读主要依赖于其针对特定传播意图所进行的即时演化与语境重构。目前，我们已公开开源 IRON-Meme 数据集及评测代码。
提交历史
来自：Gabriella Bollici [
]
[v1] 2026 年 10 月 22 日 星期四 10:51:07 UTC (7,882 KB)
English / 原文
[Submitted on 22 Oct 2026]
Title:IRON-Meme: A Benchmark for Structured Understanding of Identities and Roles in Online Meme Narratives
View a PDF of the paper titled IRON-Meme: A Benchmark for Structured Understanding of Identities and Roles in Online Meme Narratives, by Gabriella Bollici and 3 other authors
Abstract:As polarization grows on social media, understanding how political memes shape online narratives has become increasingly important. Previous research has relied on either qualitative frameworks that are hard to scale or computational methods that lack theoretical grounding. Both approaches treat narrative categories in isolation and focus exclusively on moral portrayals of identity. To address these limitations, we introduce a framework grounded in social science theories, encompassing three categories: social identities (such as gender, politics, and ethnicity), viewpoints (ingroup and outgroup perspectives), and narrative roles (moral character, intelligence, status, and power). Building on this framework, we present IRON-Meme, a human-annotated benchmark comprising 1,110 political memes across substantially more templates than existing benchmarks. Our analysis of IRON-Meme shows that most memes assign derogatory roles and that identities, roles, and templates are systematically interconnected. Evaluation of nine vision-language models indicates that frontier-scale and reasoning models achieve the highest performance, while accuracy on social identities and narrative roles decreases for less frequent labels. Incorporating template background knowledge does not enhance narrative role classification, suggesting that meme interpretation relies on how memes are adapted for specific communicative purposes. We make IRON-Meme and our code publicly available.
Submission history
From: Gabriella Bollici [
]
[v1]
Thu, 22 Oct 2026 10:51:07 UTC (7,882 KB)