# Training-Free and Interpretable Hateful Video Detection via Multi-stage Adversarial Reasoning

arXiv: https://arxiv.org/abs/2601.15115 · HTML: https://mlsanity.com/paper/2601.15115/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 25 references that are in the map.
- First posted on arXiv 21 Jan 2026.

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

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [HateXplain: A Benchmark Dataset for Explainable Hate Speech Detection](https://mlsanity.com/paper/2012.10289/) (2020)
- [Chain-of-Thought Reasoning In The Wild Is Not Always Faithful](https://mlsanity.com/paper/2503.08679/) (2025)
- [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)
- [Cross-Modal Transfer from Memes to Videos: Addressing Data Scarcity in Hateful Video Detection](https://mlsanity.com/paper/2501.15438/) (2025)
- [Towards a Robust Framework for Multimodal Hate Detection: A Study on Video vs. Image-based Content](https://mlsanity.com/paper/2502.07138/) (2025)
- [Enhanced Multimodal Hate Video Detection via Channel-wise and Modality-wise Fusion](https://mlsanity.com/paper/2505.12051/) (2025)
- [Revealing Temporal Label Noise in Multimodal Hateful Video Classification](https://mlsanity.com/paper/2508.04900/) (2025)
- [Attentive Fusion: A Transformer-based Approach to Multimodal Hate Speech Detection](https://mlsanity.com/paper/2401.10653/) (2024)

## Abstract (arXiv)

Hateful videos pose serious risks by amplifying discrimination, inciting violence, and undermining online safety. Existing training-based hateful video detection methods are constrained by limited training data and lack of interpretability, while directly prompting large vision-language models often struggle to deliver reliable hate detection. To address these challenges, this paper introduces MARS, a training-free Multi-stage Adversarial ReaSoning framework that enables reliable and interpretable hateful content detection. MARS begins with the objective description of video content, establishing a neutral foundation for subsequent analysis. Building on this, it develops evidence-based reasoning that supports potential hateful interpretations, while in parallel incorporating counter-evidence reasoning to capture plausible non-hateful perspectives. Finally, these perspectives are synthesized into a conclusive and explainable decision. Extensive evaluation on two real-world datasets shows that MARS achieves up to 10% improvement under certain backbones and settings compared to other training-free approaches and outperforms state-of-the-art training-based methods on one dataset. In addition, MARS produces human-understandable justifications, thereby supporting compliance oversight and enhancing the transparency of content moderation workflows. The code is available at https://github.com/Multimodal-Intelligence-Lab-MIL/MARS.