aiqbench analysis
This research introduces STEMTOX, a novel entropy-guided multi-task learning framework designed to improve the detection of toxic content within multimodal memes. By leveraging collaborative tags alongside visual and textual inputs, the method significantly boosts the performance of state-of-the-art Vision-Language Models (VLMs) in fine-grained toxicity classification tasks. The work addresses critical data accessibility challenges by releasing a new dataset that supports more robust online moderation systems.
STEMTOX: From Collaborative Tags to Fine-Grained Toxic Meme Detection via Entropy-Guided Multi-Task Learning
arXiv:2508.04166v3 Announce Type: replace-cross Abstract: Memes, as a widely used mode of online communication, often serve as vehicles for spreading harmful content. However, limitations in data accessibility and the high costs of dataset curation hinder the development of robust meme moderation systems. To address this challenge, in this work, we introduce a first-of-its-kind dataset - TOXICTAGS consisting of 6,300 real-world meme-based posts annotated in two stages: (i) binary classification into toxic and normal, and (ii) fine-grained labelling of toxic memes as hateful, dangerous, or offensive. A key feature of this dataset is that it includes collaborative tags associated with the original posts, enhancing the context of each meme. In addition, we propose a novel entropy-guided multi-tasking framework -- STEMTOX -- that leverages these collaborative tags alongside visual and textual inputs within a robust classification framework. Experimental results show that incorporating these tags substantially enhances the performance of state-of-the-art VLMs in toxicity detection tasks. Our contributions offer a novel and scalable foundation for improved content moderation in multimodal online environments. Warning: Contains potentially toxic contents.
Abstract:Memes, as a widely used mode of online communication, often serve as vehicles for spreading harmful content. However, limitations in data accessibility and the high costs of dataset curation hinder the development of robust meme moderation systems. To address this challenge, in this work, we introduce a first-of-its-kind dataset - TOXICTAGS consisting of 6,300 real-world meme-based posts annotated in two stages: (i) binary classification into toxic and normal, and (ii) fine-grained labelling of toxic memes as hateful, dangerous, or offensive. A key feature of this dataset is that it includes collaborative tags associated with the original posts, enhancing the context of each meme. In addition, we propose a novel entropy-guided multi-tasking framework -- STEMTOX -- that leverages these collaborative tags alongside visual and textual inputs within a robust classification framework. Experimental results show that incorporating these tags substantially enhances the performance of state-of-the-art VLMs in toxicity detection tasks. Our contributions offer a novel and scalable foundation for improved content moderation in multimodal online environments. Warning: Contains potentially toxic contents.
Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
arXiv:2508.04166 [cs.CV]
(or arXiv:2508.04166v3 [cs.CV] for this version)
Full text under the publisher's open licence · 187 words
Related coverage
AgentOps: Operationalize agentic AI at scale with Amazon Bedrock AgentCore
AWS · Jun 1, 2026
SAFE-MEME: Structured Reasoning Framework for Robust Hate Speech Detection in Memes
arXiv · Jul 28, 2026
AHA-Memes: A Fine-Grained Multimodal Benchmark for Understanding Hate in Arabic Memes
arXiv · Jul 31, 2026
Pad\=artha: Ontology-Grounded Fine-Grained NER Benchmark for Classical Sanskrit
arXiv · Sep 1, 2026