# Understanding and Mitigating Toxicity in Image-Text Pretraining Datasets: A Case Study on LLaVA

arXiv: https://arxiv.org/abs/2505.06356 · HTML: https://mlsanity.com/paper/2505.06356/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 29 references that are in the map.
- First posted on arXiv 9 May 2025.

## Builds on (most cited first, 20 of 20)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [Kosmos-2: Grounding Multimodal Large Language Models to the World](https://mlsanity.com/paper/2306.14824/) (2023)
- [PaLI: A Jointly-Scaled Multilingual Language-Image Model](https://mlsanity.com/paper/2209.06794/) (2022)
- [Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks](https://mlsanity.com/paper/2311.06242/) (2023)
- [MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models](https://mlsanity.com/paper/2311.17600/) (2023)
- [PaLI-X: On Scaling up a Multilingual Vision and Language Model](https://mlsanity.com/paper/2305.18565/) (2023)
- [Kosmos-G: Generating Images in Context with Multimodal Large Language Models](https://mlsanity.com/paper/2310.02992/) (2023)
- [SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model](https://mlsanity.com/paper/2406.12030/) (2024)
- [Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages](https://mlsanity.com/paper/2410.16153/) (2024)
- [From Pixels to Prose: A Large Dataset of Dense Image Captions](https://mlsanity.com/paper/2406.10328/) (2024)
- [Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection](https://mlsanity.com/paper/2402.12501/) (2024)
- [CoMM: A Coherent Interleaved Image-Text Dataset for Multimodal Understanding and Generation](https://mlsanity.com/paper/2406.10462/) (2024)
- [Understanding and Rectifying Safety Perception Distortion in VLMs](https://mlsanity.com/paper/2502.13095/) (2025)
- [Zero-Shot Defense Against Toxic Images via Inherent Multimodal Alignment in LVLMs](https://mlsanity.com/paper/2503.00037/) (2025)

## Built on by (most cited first, 1 of 1)

- [TADS: Task-Aware Data Selection for Multi-Task Multimodal Pre-Training](https://mlsanity.com/paper/2602.05251/) (2026)

## Abstract (arXiv)

Pretraining datasets are foundational to the development of multimodal models, yet they often have inherent biases and toxic content from the web-scale corpora they are sourced from. In this paper, we investigate the prevalence of toxicity in LLaVA image-text pretraining dataset, examining how harmful content manifests in different modalities. We present a comprehensive analysis of common toxicity categories and propose targeted mitigation strategies, resulting in the creation of a refined toxicity-mitigated dataset. This dataset removes 7,531 of toxic image-text pairs in the LLaVA pre-training dataset. We offer guidelines for implementing robust toxicity detection pipelines. Our findings underscore the need to actively identify and filter toxic content - such as hate speech, explicit imagery, and targeted harassment - to build more responsible and equitable multimodal systems. The toxicity-mitigated dataset is open source and is available for further research.