# SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset

arXiv: https://arxiv.org/abs/2406.14477 · HTML: https://mlsanity.com/paper/2406.14477/

## Key facts

- Cited by 35 papers in the map (data through 14 Oct 2026).
- Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 41 of its 95 references that are in the map.
- First posted on arXiv 20 Jun 2024.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Diffusion and multimodal model alignment](https://mlsanity.com/alignment/diffusion-and-multimodal-model-alignment/)

## Builds on (most cited first, 30 of 41)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs](https://mlsanity.com/paper/2111.02114/) (2021)
- [Video-LLaVA: Learning United Visual Representation by Alignment Before Projection](https://mlsanity.com/paper/2311.10122/) (2023)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [Scaling Laws for Reward Model Overoptimization](https://mlsanity.com/paper/2210.10760/) (2022)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity](https://mlsanity.com/paper/2302.04023/) (2023)
- [BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset](https://mlsanity.com/paper/2307.04657/) (2023)
- [Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned](https://mlsanity.com/paper/2209.07858/) (2022)
- [ModelScope Text-to-Video Technical Report](https://mlsanity.com/paper/2308.06571/) (2023)
- [Safe RLHF: Safe Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2310.12773/) (2023)
- [VideoCrafter1: Open Diffusion Models for High-Quality Video Generation](https://mlsanity.com/paper/2310.19512/) (2023)
- [InternVid: A Large-scale Video-Text Dataset for Multimodal Understanding and Generation](https://mlsanity.com/paper/2307.06942/) (2023)
- [Tutorial on Variational Autoencoders](https://mlsanity.com/paper/1606.05908/) (2016)
- [Latent Video Diffusion Models for High-Fidelity Long Video Generation](https://mlsanity.com/paper/2211.13221/) (2022)
- [Generating Videos with Scene Dynamics](https://mlsanity.com/paper/1609.02612/) (2016)
- [LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment](https://mlsanity.com/paper/2310.01852/) (2023)
- [EvalCrafter: Benchmarking and Evaluating Large Video Generation Models](https://mlsanity.com/paper/2310.11440/) (2023)
- [Unsolved Problems in ML Safety](https://mlsanity.com/paper/2109.13916/) (2021)
- [Model evaluation for extreme risks](https://mlsanity.com/paper/2305.15324/) (2023)
- [Swap Attention in Spatiotemporal Diffusions for Text-to-Video Generation](https://mlsanity.com/paper/2305.10874/) (2023)
- [Jailbreaking Attack against Multimodal Large Language Model](https://mlsanity.com/paper/2402.02309/) (2024)

## Built on by (most cited first, 30 of 35)

- [SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generation](https://mlsanity.com/paper/2410.12761/) (2024)
- [Improving Dynamic Object Interactions in Text-to-Video Generation with AI Feedback](https://mlsanity.com/paper/2412.02617/) (2024)
- [Align Anything: Training All-Modality Models to Follow Instructions with Language Feedback](https://mlsanity.com/paper/2412.15838/) (2024)
- [Video-SafetyBench: A Benchmark for Safety Evaluation of Video LVLMs](https://mlsanity.com/paper/2505.11842/) (2025)
- [NarrLV: Towards a Comprehensive Narrative-Centric Evaluation for Long Video Generation](https://mlsanity.com/paper/2507.11245/) (2025)
- [T2I-RiskyPrompt: A Benchmark for Safety Evaluation, Attack, and Defense on Text-to-Image Model](https://mlsanity.com/paper/2510.22300/) (2025)
- [MJ-VIDEO: Fine-Grained Benchmarking and Rewarding Video Preferences in Video Generation](https://mlsanity.com/paper/2502.01719/) (2025)
- [T2VUnlearning: A Concept Erasing Method for Text-to-Video Diffusion Models](https://mlsanity.com/paper/2505.17550/) (2025)
- [VideoEraser: Concept Erasure in Text-to-Video Diffusion Models](https://mlsanity.com/paper/2508.15314/) (2025)
- [Position: Interactive Generative Video as Next-Generation Game Engine](https://mlsanity.com/paper/2503.17359/) (2025)
- [Video Unlearning via Low-Rank Refusal Vector](https://mlsanity.com/paper/2506.07891/) (2025)
- [Can You Count to Nine? A Human Evaluation Benchmark for Counting Limits in Modern Text-to-Video Models](https://mlsanity.com/paper/2504.04051/) (2025)
- [GRADEO: Towards Human-Like Evaluation for Text-to-Video Generation via Multi-Step Reasoning](https://mlsanity.com/paper/2503.02341/) (2025)
- [Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks](https://mlsanity.com/paper/2607.07907/) (2026)
- [Multi-level Value Alignment in Agentic AI Systems: Survey and Perspectives](https://mlsanity.com/paper/2506.09656/) (2025)
- [PurifyGen: A Risk-Discrimination and Semantic-Purification Model for Safe Text-to-Image Generation](https://mlsanity.com/paper/2512.23546/) (2025)
- [InterMT: Multi-Turn Interleaved Preference Alignment with Human Feedback](https://mlsanity.com/paper/2505.23950/) (2025)
- [Now You See It, Now You Don't - Instant Concept Erasure for Safe Text-to-Image and Video Generation](https://mlsanity.com/paper/2511.18684/) (2025)
- [CGCE: Classifier-Guided Concept Erasure in Generative Models](https://mlsanity.com/paper/2511.05865/) (2025)
- [SafeGen-Bench: Benchmarking Safety in Image-Conditioned Text-to-Video Generation](https://mlsanity.com/paper/2606.01481/) (2026)
- [Activation Steering of Video Generation Models via Reduced-Order Linear Optimal Control](https://mlsanity.com/paper/2606.04775/) (2026)
- [GuardReasoner-Omni: A Reasoning-based Multi-modal Guardrail for Text, Image, Video, and Audio](https://mlsanity.com/paper/2602.03328/) (2026)
- [AVFakeBench: A Comprehensive Audio-Video Forgery Detection Benchmark for AV-LMMs](https://mlsanity.com/paper/2511.21251/) (2025)
- [Pulling The REINS: Training-Free Safety Alignment of Video Diffusion Models via Representation Steering](https://mlsanity.com/paper/2606.17257/) (2026)
- [On Monotonicity in AI Alignment](https://mlsanity.com/paper/2506.08998/) (2025)
- [HKGAI-V1: Towards Regional Sovereign Large Language Model for Hong Kong](https://mlsanity.com/paper/2507.11502/) (2025)
- [DesignSense: A Human Preference Dataset and Reward Modeling Framework for Graphic Layout Generation](https://mlsanity.com/paper/2602.23438/) (2026)
- [Prompt-to-Gesture: Measuring the Capabilities of Image-to-Video Deictic Gesture Generation](https://mlsanity.com/paper/2604.14953/) (2026)
- [PACT: Self-Evolving Physical Safety Alignment for Diffusion Policies in Embodied Manipulation](https://mlsanity.com/paper/2606.08414/) (2026)
- [Uni-AdaVD: Universal Concept Erasure for Visual Generation via Orthogonal Value Decomposition](https://mlsanity.com/paper/2607.14521/) (2026)

## Abstract (arXiv)

To mitigate the risk of harmful outputs from large vision models (LVMs), we introduce the SafeSora dataset to promote research on aligning text-to-video generation with human values. This dataset encompasses human preferences in text-to-video generation tasks along two primary dimensions: helpfulness and harmlessness. To capture in-depth human preferences and facilitate structured reasoning by crowdworkers, we subdivide helpfulness into 4 sub-dimensions and harmlessness into 12 sub-categories, serving as the basis for pilot annotations. The SafeSora dataset includes 14,711 unique prompts, 57,333 unique videos generated by 4 distinct LVMs, and 51,691 pairs of preference annotations labeled by humans. We further demonstrate the utility of the SafeSora dataset through several applications, including training the text-video moderation model and aligning LVMs with human preference by fine-tuning a prompt augmentation module or the diffusion model. These applications highlight its potential as the foundation for text-to-video alignment research, such as human preference modeling and the development and validation of alignment algorithms.