# TANDEM: Temporal-Aware Neural Detection for Multimodal Hate Speech

arXiv: https://arxiv.org/abs/2601.11178 · HTML: https://mlsanity.com/paper/2601.11178/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 53 references that are in the map.
- First posted on arXiv 16 Jan 2026.

## Builds on (most cited first, 12 of 12)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [Temporal Action Detection with Structured Segment Networks](https://mlsanity.com/paper/1704.06228/) (2017)
- [Self-Rewarding Vision-Language Model via Reasoning Decomposition](https://mlsanity.com/paper/2508.19652/) (2025)
- [TRACE: Temporal Grounding Video LLM via Causal Event Modeling](https://mlsanity.com/paper/2410.05643/) (2024)
- [On the Origins of Memes by Means of Fringe Web Communities](https://mlsanity.com/paper/1805.12512/) (2018)
- [HateMM: A Multi-Modal Dataset for Hate Video Classification](https://mlsanity.com/paper/2305.03915/) (2023)
- [MultiHateClip: A Multilingual Benchmark Dataset for Hateful Video Detection on YouTube and Bilibili](https://mlsanity.com/paper/2408.03468/) (2024)
- [ImpliHateVid: A Benchmark Dataset and Two-stage Contrastive Learning Framework for Implicit Hate Speech Detection in Videos](https://mlsanity.com/paper/2508.06570/) (2025)
- [Towards a Robust Framework for Multimodal Hate Detection: A Study on Video vs. Image-based Content](https://mlsanity.com/paper/2502.07138/) (2025)
- [MultiHateLoc: Towards Temporal Localisation of Multimodal Hate Content in Online Videos](https://mlsanity.com/paper/2512.10408/) (2025)

## Abstract (arXiv)

Social media platforms are increasingly dominated by long-form multimodal content, where harmful narratives are constructed through a complex interplay of audio, visual, and textual cues. While automated systems can flag hate speech with high accuracy, they often function as "black boxes" that fail to provide the granular, interpretable evidence, such as precise timestamps and target identities, required for effective human-in-the-loop moderation. In this work, we introduce TANDEM, a unified framework that transforms audio-visual hate detection from a binary classification task into a structured reasoning problem. Our approach employs a novel tandem reinforcement learning strategy where vision-language and audio-language models optimize each other through self-constrained cross-modal context, stabilizing reasoning over extended temporal sequences without requiring dense frame-level supervision. Experiments across three benchmark datasets demonstrate that TANDEM significantly outperforms zero-shot and context-augmented baselines, achieving 0.73 F1 in target identification on HateMM (a 30% improvement over state-of-the-art) while maintaining precise temporal grounding. We further observe that while binary detection is robust, differentiating between offensive and hateful content remains challenging in multi-class settings due to inherent label ambiguity and dataset imbalance. More broadly, our findings suggest that structured, interpretable alignment is achievable even in complex multimodal settings, offering a blueprint for the next generation of transparent and actionable online safety moderation tools.