# UNIVID: Unified Vision-Language Model for Video Moderation

arXiv: https://arxiv.org/abs/2606.05748 · HTML: https://mlsanity.com/paper/2606.05748/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 19 of its 33 references that are in the map.
- First posted on arXiv 4 Jun 2026.

## Placed in

- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)

## Builds on (most cited first, 19 of 19)

- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Video-LLaVA: Learning United Visual Representation by Alignment Before Projection](https://mlsanity.com/paper/2311.10122/) (2023)
- [OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge](https://mlsanity.com/paper/1906.00067/) (2019)
- [EVA-CLIP: Improved Training Techniques for CLIP at Scale](https://mlsanity.com/paper/2303.15389/) (2023)
- [Real-world Anomaly Detection in Surveillance Videos](https://mlsanity.com/paper/1801.04264/) (2018)
- [Perception Test: A Diagnostic Benchmark for Multimodal Video Models](https://mlsanity.com/paper/2305.13786/) (2023)
- [LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training](https://mlsanity.com/paper/2509.23661/) (2025)
- [Tarsier: Recipes for Training and Evaluating Large Video Description Models](https://mlsanity.com/paper/2407.00634/) (2024)
- [Human-AI Collaboration via Conditional Delegation: A Case Study of Content Moderation](https://mlsanity.com/paper/2204.11788/) (2022)
- [Video-SafetyBench: A Benchmark for Safety Evaluation of Video LVLMs](https://mlsanity.com/paper/2505.11842/) (2025)
- [VLM as Policy: Common-Law Content Moderation Framework for Short Video Platform](https://mlsanity.com/paper/2504.14904/) (2025)
- [Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Governance](https://mlsanity.com/paper/2509.21486/) (2025)
- [AI vs. Human Moderators: A Comparative Evaluation of Multimodal LLMs in Content Moderation for Brand Safety](https://mlsanity.com/paper/2508.05527/) (2025)
- [Validating Multimedia Content Moderation Software via Semantic Fusion](https://mlsanity.com/paper/2305.13623/) (2023)
- [CPFD: Confidence-aware Privileged Feature Distillation for Short Video Classification](https://mlsanity.com/paper/2410.03038/) (2024)
- [Embedding-based Retrieval in Multimodal Content Moderation](https://mlsanity.com/paper/2507.01066/) (2025)
- [From Native Memes to Global Moderation: Cross-Cultural Evaluation of Vision-Language Models for Hateful Meme Detection](https://mlsanity.com/paper/2602.07497/) (2026)

## Abstract (arXiv)

Global-scale video moderation faces a dual challenge: the need for fine-grained multi-modal reasoning and the demand for interpretable outputs to support downstream enforcement. Traditional moderation systems often rely on fragmented black-box classifiers that are difficult to maintain and lack transparency. In this paper, we present UNIVID, a UNIfied VIsion-language model for video moDeration. Unlike standard classification models, UNIVID generates policy-aware captions that serve as an interpretable intermediate representation, enabling human-verifiable decisions and multi-task reusability. While existing open-source and commercial VLMs often suffer from safety-guardrail refusals and lack fine-grained policy alignment, we develop a specialized training data recipe that combines expert human-refined labels with synthetic data to align the model with our safety guidelines. By integrating UNIVID as the core captioner, we design a novel end-to-end video moderation system that reduces violation leakage by 42.7% and overkill rate by 37.0% relatively. Meanwhile, by replacing over 1,000 policy-specific models with a single UNIVID backbone, we recycled extensive computation resources while reducing engineering maintenance overhead. To our knowledge, this is one of the first reports of a high-efficiency captioning VLM successfully supporting industrial-scale moderation and cross-functional business.