# Learning Human-Perceived Fakeness in AI-Generated Videos via Multimodal LLMs

arXiv: https://arxiv.org/abs/2509.22646 · HTML: https://mlsanity.com/paper/2509.22646/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Placed in Video understanding with MLLMs, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 32 of its 48 references that are in the map.
- First posted on arXiv 26 Sep 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)

## Builds on (most cited first, 30 of 32)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone](https://mlsanity.com/paper/2404.14219/) (2024)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [Towards Accurate Generative Models of Video: A New Metric & Challenges](https://mlsanity.com/paper/1812.01717/) (2018)
- [MVBench: A Comprehensive Multi-modal Video Understanding Benchmark](https://mlsanity.com/paper/2311.17005/) (2023)
- [Gemini: A Family of Highly Capable Multimodal Models](https://mlsanity.com/paper/2312.11805/) (2023)
- [Phi-4 Technical Report](https://mlsanity.com/paper/2412.08905/) (2024)
- [LLaVA-Video: Video Instruction Tuning With Synthetic Data](https://mlsanity.com/paper/2410.02713/) (2024)
- [VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding](https://mlsanity.com/paper/2501.13106/) (2025)
- [DIRE for Diffusion-Generated Image Detection](https://mlsanity.com/paper/2303.09295/) (2023)
- [LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion Models](https://mlsanity.com/paper/2309.15103/) (2023)
- [Improving Video Generation with Human Feedback](https://mlsanity.com/paper/2501.13918/) (2025)
- [Automatic Detection of Generated Text is Easiest when Humans are Fooled](https://mlsanity.com/paper/1911.00650/) (2019)
- [VideoPhy: Evaluating Physical Commonsense for Video Generation](https://mlsanity.com/paper/2406.03520/) (2024)
- [A Note on the Inception Score](https://mlsanity.com/paper/1801.01973/) (2018)
- [GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium](https://mlsanity.com/paper/1706.08500/) (2017)
- [RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors](https://mlsanity.com/paper/2405.07940/) (2024)
- [Hierarchical Fine-Grained Image Forgery Detection and Localization](https://mlsanity.com/paper/2303.17111/) (2023)
- [TALL: Thumbnail Layout for Deepfake Video Detection](https://mlsanity.com/paper/2307.07494/) (2023)
- [DeMamba: AI-Generated Video Detection on Million-Scale GenVideo Benchmark](https://mlsanity.com/paper/2405.19707/) (2024)
- [Spatiotemporal Inconsistency Learning for DeepFake Video Detection](https://mlsanity.com/paper/2109.01860/) (2021)
- [T2V-Turbo-v2: Enhancing Video Generation Model Post-Training through Data, Reward, and Conditional Guidance Design](https://mlsanity.com/paper/2410.05677/) (2024)
- [Generalizable Synthetic Image Detection via Language-guided Contrastive Learning](https://mlsanity.com/paper/2305.13800/) (2023)
- [VBench++: Comprehensive and Versatile Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2411.13503/) (2024)

## Built on by (most cited first, 17 of 17)

- [Skyra: AI-Generated Video Detection via Grounded Artifact Reasoning](https://mlsanity.com/paper/2512.15693/) (2025)
- [VideoVeritas: AI-Generated Video Detection via Perception Pretext Reinforcement Learning](https://mlsanity.com/paper/2602.08828/) (2026)
- [Can Textual Reasoning Improve the Performance of MLLMs on Fine-grained Visual Classification?](https://mlsanity.com/paper/2601.06993/) (2026)
- [Preserving Forgery Artifacts: AI-Generated Video Detection at Native Scale](https://mlsanity.com/paper/2604.04634/) (2026)
- [ViBES: A Conversational Agent with Behaviorally-Intelligent 3D Virtual Body](https://mlsanity.com/paper/2512.14234/) (2025)
- [FusionAgent: A Multimodal Agent with Dynamic Model Selection for Human Recognition](https://mlsanity.com/paper/2603.26908/) (2026)
- [What's Missing in Screen-to-Action? Towards a UI-in-the-Loop Paradigm for Multimodal GUI Reasoning](https://mlsanity.com/paper/2604.06995/) (2026)
- [When Images Look Right and Retrieve Wrong: Coverage-Guided Cross-Scale Re-Indexing for Knowledge-Faithful Generative Perception](https://mlsanity.com/paper/2608.20810/) (2026)
- [Self-Evaluation Is Already There: Eliciting Latent Judge Calibration in Base LLMs with Minimal Data](https://mlsanity.com/paper/2606.05122/) (2026)
- [AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoning](https://mlsanity.com/paper/2608.29988/) (2026)
- [Video as Natural Augmentation: Towards Unified AI-Generated Image and Video Detection](https://mlsanity.com/paper/2605.21977/) (2026)
- [Detecting AI-Generated Video: A Vision-Language Dual-View Survey](https://mlsanity.com/paper/2607.10787/) (2026)
- [VidForensics-M1: Meta-Detection Reinforcement Learning with Verifiable Temporal Grounding for AI-Generated Video Forensics](https://mlsanity.com/paper/2608.11201/) (2026)
- [EA-Swin: An Embedding-Agnostic Swin Transformer for AI-Generated Video Detection](https://mlsanity.com/paper/2602.17260/) (2026)
- [VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis](https://mlsanity.com/paper/2605.22570/) (2026)
- [Noise Floor Audit for Agent Benchmarks](https://mlsanity.com/paper/2608.22331/) (2026)
- [When LLM Meets Tree Search: A Systematic View of Inference as Search in Large Language Models](https://mlsanity.com/paper/2608.30395/) (2026)

## Abstract (arXiv)

Can humans identify AI-generated (fake) videos and provide grounded reasons? While video generation models have advanced rapidly, a critical dimension -- whether humans can detect deepfake traces within a generated video, i.e., spatiotemporal grounded visual artifacts that reveal a video as machine generated -- has been largely overlooked. We introduce DeeptraceReward, the first fine-grained, spatially- and temporally- aware benchmark that annotates human-perceived fake traces for video generation reward. The dataset comprises 4.3K detailed annotations across 3.3K high-quality generated videos. Each annotation provides a natural-language explanation, pinpoints a bounding-box region containing the perceived trace, and marks precise onset and offset timestamps. We consolidate these annotations into 9 major categories of deepfake traces that lead humans to identify a video as AI-generated, and train multimodal language models (LMs) as reward models to mimic human judgments and localizations. On DeeptraceReward, our 7B reward model outperforms GPT-5 by 34.7% on average across fake clue identification, grounding, and explanation. Interestingly, we observe a consistent difficulty gradient: binary fake v.s. real classification is substantially easier than fine-grained deepfake trace detection; within the latter, performance degrades from natural language explanations (easiest), to spatial grounding, to temporal labeling (hardest). By foregrounding human-perceived deepfake traces, DeeptraceReward provides a rigorous testbed and training signal for socially aware and trustworthy video generation.